SoarNeck
Dec 21, 2003, 09:35 PM
!!The Wing
While the wing is originally designed to be a single-piece unit, I wanted a two-piece wing to make transporting and storing it a lot easier. I ended up using a 5/8" solid aluminum rod as my joiner rod, since I have a lot of it on hand for when I build sailplane wings. The tube that sits in the wing is a 5/8" inner diameter carbon tube, available from Aerospace Composites (ACP) or The Composite Store (CST).
The finished wing is taped together before flight, and held on with the kit-specified rubber bands. I generally prefer a bolt-on wing arrangement, but I thought that bands would make quick battery changes a little easier. Do remember to use enough rubber bands to keep the wing secured.
Since I didn't have any angled rod in stock and the Telemaster wing has a fair amount of dihedral in it, I ended up using a fairly short wing joiner tube. Given that I ended up having problems with wing strength and that the dihedral isn't required for a tug anyway, if I built another Telemaster I would remove a fair amount of dihedral and use a longer joiner tube (maybe three bays per wing instead of only two). The joiner rod itself never bent at all.
I also changed the shear web arrangement from the kit spec, and used vertical grain balsa between the spars.Webs glued to the face of the spar are inefficient, since they really only test shear strength of the glue and the balsa to the depth that the glue penetrates. The first third of the wing used balsa the full width of the spar, and then I gradually tapered the thickness of the web towards the tip, which used only 1/16" balsa webs. Be sure to keep the webs centered in the middle of the spar. Finally, I wrapped the spar from the root to one bay beyond the end of the joiner tube with Kevlar thread, available from ACP or CST. This helps to prevent the spar from bursting at the ends of the tube, where such stresses are typically the greatest.
@905504:Since the joiner rod is straight, the joiner tube must sit in the wing at an angle.
I don't like torque rods of any description, so I decided to mount the servos in the wing (Multiplex MCV2 digitals, which were on hand). Mounting them was a simple affair, consisting of a couple of balsa braces. I decided to leave the servos exposed to the airstream instead of concealed in the wing, since this is a powered model after all.
@905505:Blocking the aileron servo in place.
@905506:A small triangular patch of balsa used to lead out the aileron extension. This ended up being too close to the outside of the fuselage, and so the wire was eventually led out against the root rib. Note the kevlar spar wrapping.
!!Final Assembly
The finished structure is light and efficient when built stock, so I didn't go to many great lengths to lighten the structure. I cut lightening holes in the tail ribs with only a very slight benefit in weight, and gave up after lightening a few wing ribs simply because the balsa was hard enough that the Exacto knife hurt my fingers after a while! (I used them anyway)All of this seems like sacrilege given what we've learned about electric flight, but remember that a heavier tug actually works to our advantage in the air, since a heavy tug tracks better on tow. Also try to remember that saving even 4 ounces on a 200+ oz model is of minimal benefit, especially if it takes hours to cut new ribs or lightening holes.
@905507:The finished model "in the bones", on the stock aluminum gear. Getting to this stage took me about 5 sessions in the shop, as building proceeds quickly.
I chose to cover the model in various colours of Ultracote, which took almost as long to do as building the airplane in the first place (those panel lines were all cut from self-adhesive trim sheet). I like Ultracote for its ease of application, and because it doesn't get as brittle as Monokote as it ages (or in low temperatures).
While it would have been easier to do so earlier, after covering the airplane I sheeted the majority of the batter compartment with 1/32" ply. There really isn't much wood available to support the internal components mounted there, and the fuselage needed a bit of stiffening in that area.
Next up was the pack, which I assembled using six sticks of five cells soldered end to end. The wrappers on GMVIS cells can be tricky to remove, but with patience it can be done cleanly and reasonably quickly (hint: try not to tear the wrapper, or it will shred and leave remnants on the cell that are hard to remove without Goo-Gone). Each stick was then wrapped in 1 mil Kapton tape individually, and the sticks were banded together with another wrap of the tape. This treatment is usually reserved for extreme current applications, but I find that all cells benefit from the increased cooling that Kapton allows, and Kapton is much safer since it never melts (within reason; if the pack gets to 500F, the tape melting isn't your biggest problem).
Mounting the battery is of utmost importance on a big model, since a 30 cell pack weighs just under four pounds and can carry a lot of momentum. Given the cost of a big pack for this airplane, I elected to make life easy by not making the pack removable, and blocked it in place with balsa sticks reinforced with thin ply and balsa gussets. It wouldn't be hard to do so, however. One system would be to use filament tape (or Kapton tape for high-current tugs) to attach angle brackets made of ply or plastic to the battery pack, and to retain the pack by screwing these angle brackets to matching counterparts in the fuselage. That system would also allow easy changes of c/g, if the fuselage track had a series of matching bolt holes.
@905508:The pack mounted in the airplane. Balsa gussets were added to the ply battery hold-downs after I started flying the model aerobatically, mainly as a precaution.
I use 4mm gold pins for my pack connections, since I have them on hand from high-current projects. The Sermos connector pictured below merely adapts the pack to my charger leads.
@905509:The Aveox 1817/4Y motor fitted to one of Tom Hunt's ModelAirTech H1500 belt drives. The pulleys here represent the earlier 2:1 ratio.
@905510:A scoop is usually fitted over the hole in the bottom of the fuselage, but it had been temporarily removed here for access to the speed control wires.Bare_Bones_Stock_Gear.jpg"></A>
Since the Aveox speed control is an opto unit, I used a receiver pack made up from five 1100mah Sanyo AA cells. The receiver is a Multiplex IPD 9 channel model, and I fly with a Multiplex Profi 3030 transmitter.
!Flying
After the tug was finally completed, a couple of run-up tests were performed to verify that the drive system was working properly. Unfortunately, the driveline started to make a horrible vibrating sound at about half throttle, which made me nervous about running it any harder. At this point the belt drive was setup as recommended, with the motor lightly contacting the four stabilizing setscrews and a single zip-tie maintaining the contact. After painstakingly balancing the 22x10 Top Flite Powerpoint prop (no difference), I was getting discouraged. A quick email to Tom Hunt suggested that the prop might not be tracking properly, and since I was able to see that the blades weren't perfectly tracking up, I went back to my local hobby store to find another prop.
@905511:The Telemaster is a good looking design, and I'm especially happy with the way the covering job turned out.
I came home with a Zinger 20x11 wood prop, and again balanced it as best I could. No difference, the vibration was still there. I could feel a bit of play between the output shaft on the belt drive and the collet-style prop adaptor, so I asked Tom to send me a new adaptor (which he did without hesitation). Tom mentioned that he had never had any problems with these drives, and that they should operate smoothly throughout the whole RPM range as long as the motor spun at less than 20,000 revs (we were only spinning 10,000), so I had hope that we would be able to lick the vibration issue. When the new adaptor arrived, the vibration was a little lessened, but still not gone. Thinking that a wrap of tape around the output shaft (under the collet) might help, I added a single layer of Kapton tape under the adaptor, and that seemed to get rid of most of the problem. I also ordered some of the new 20x13" APC-E props, thinking that a molded plastic prop would be more likely to run true than a wooden version (which might have warped due to the low humidity that Calgary is famous for).
@905512:The view from behind shows how tall the gear is. The model didn't roll very well through rough grass on the original 3" Robart tires shown here, and they were later changed to 4.5" Sullivan Featherlites.
At that point, I decided to get another set of eyes on the vibration issue, and so invited Ryan over to give his opinion. After much testing and debate, Ryan noticed that the motor still seemed to be vibrating against the stabilizing setscrews. On his suggestion, we added two zip ties in place of the original single one, and another pair around the end of the motor (for four in total). This allowed us to run the motor up to full power with either prop, and the model looked to be ready for a test flight.
The first flights took place at a local pylon racing field, which had a nice concrete runway that we could use. Thanks again to the Calgary<A href="http://www.cmpra.com/"> Miniature Pylon Racing Association </A>for use of their field! The wind was blowing at about 25kph gusting to 35 or so, which was not exactly calm, but I was confident that the 13 lb model could handle the wind. Taxiing was impossible on the high gear and small wheels, but Ryan kindly offered to hold the model in place until I was ready for takeoff.After slowly advancing the throttle to half, the Telemaster was off the ground in about 5 feet! At that point it was climbing strongly, but not making much progress forward. I could tell that the model was going to be a slow flier on the 22x10 prop, but hoped that it would okay in less wind.
The rest of the 4 min flight was largely uneventful, since the wind didn't allow for much beyond a couple of circuits on the field. Landing was easy given the low groundspeed, but I held a bit more speed than normal into the flair to prevent the model from ground looping. Even with that precaution, I was thankful that Ryan was there to catch the wingtip at the end of the flight, since the Telemaster tried to blow over after coming to a stop! That high gear is nice for prop clearance, but can make the airplane a bit tippy on the group.
@905513:The Telemaster coming in on final at the CMPRA field.
@905514:I really liked the look of the model in the air, and it seemed stable even in very active air.
Since the tug seemed to be ready to go, it was time to see how the model behaved with a sailplane behind it. The next weekend was warm and calm (unusual for Calgary in the fall), and seemed like the perfect time to try towing. We were originally going to try towing from the local full-size sailplane field, but it was being used on that weekend for normal operations. The backup location was our own model club field, which was a little rough, but serviceable. The grass was reasonably short since it was late in the year, and while the ground was a little rough (it isn't intended to service wheeled models), the big airplanes seemed to handle it okay.
As Ryan readied his ASK for flight, I took the chance to put a couple of more orientation flights on the tug. It was obviously not lacking in power, coming off the ground in 10 feet at 1/3 throttle (!), but the controllability also seemed good for such a large model as well. There wasn't a noticeable change in pitch trim when the throttle was advanced, and the balance point seemed to be only slightly forward on the C/G. Things seemed to be good to go, so I borrowed Ryan's Schulze 630d charger to top up the pack, and went to work in putting together some tow lines.
We decided to use 200lb braided Dacron tow lines for a couple of reasons, not the least of which was that it was the same line that I use on my winch. In addition to being available, the Dacron line offered a nice balance between strength, stretchiness, and visibility. 80lb monofilament fishing line loops were used on either end of the line to allow it to be fed into the tow releases. It would be a good idea to have a swiveled fishing snap on the sailplane end of the line, since that would allow for quick attachments to loops that are pre-installed in the sailplane's release. Additionally, when the loops are released under load, the cam-style tow releases tend to nick the monofilament loops, which means that replacing them every flight isn't a bad idea. 120-150 foot lines are the norm at our field.
+905518:First Steps
After finding the patch of ground that was most level and free from badger holes, we hooked the models together and pointed them into the wind (2-5mph). The plan was to release after making one large of the field, with the first turn to the left. It's always better for the tug and sailplane pilots to agree upon a flight plan before taking off,especially when both pilots are relatively inexperienced.Additionally, be sure to release at the first sign of trouble, since things can get ugly quickly.
After taking a couple of deep breaths, it was time to go! Smoothly but quickly advancing the throttle, the combo accelerated strongly. There was no problem keeping either the sailplane or the tug tracking straight, and the climb rate seemed to be good. Unfortunately, the belt drive started making ugly sounds again (see Video 1), though they were only under the load of a sailplane.
After taking off, we proceeded to climb out in a straight line to the southwest. We had always been told to tow at full throttle, and to modulate the speed of the tow by climb angle, but this was proving difficult in practice. I was having trouble keeping the tail of the tug up, and so Ryan decided to release before things got too slow. We had managed to get to about 300 feet quite quickly, but since the sailplane was well below the tug, it seemed safer to try again after making some adjustments.
The sailplane had enough time to make a couple of passes over the lakebed, and even wandered though a promising patch of lift. However, since it was only his second flight with the airplane, Ryan wisely decided to land instead. After making a low pass to drop the towline, I dropped the tug in for a gently landing soon thereafter. While I could have elected to land with the line still attached, I feel that it's safer to drop it instead. After watching a tug snag a fenceline with its towline (and subsequently pancake into the deck really hard), I've always followed this principle.
+905519: Trying Again
After recharging the battery (it took about 500mah, so this wasn't strictly necessary), we lined the models up for another tow. I increased the amount of down elevator travel that I had available, and added some down trim to the tug as well. The second takeoff went as easily as the first, though in a more coordinated fashion. Since I have trouble watching two models at once, I had asked Ryan to tell me when his sailplane was off the ground. This prevents the dangerous situation where the tug takes off first, before the sailplane is fully up to speed. A well-behaved sailplane like the ASK-21 might not object to be dragged around barely above stall speed, but an aerobatic model like aFox surely won't appreciate it. At any rate, I "flew" the tug along the ground until I heard the call from Ryan, and there were no problems at all.
On climbout, I still found that I had to hold some down elevator, but the climbout was much more controllable in pitch than it had been before. A new problem arose with the first turn however...the tug didn't want to change course! I needed full aileron and a good deal of rudder to coordinate the turn, which I attributed both to fighting the Telemaster's inherent stability and the right moment from the tow release (this would be eliminated by towing from on the tug c/g instead of from behind it). Consequently, I inadvertently let the tug dive a little bit through the turn as I concentrated on keeping the turn large and...
+905520: Things Get Ugly
Ugly is putting it mildly. From what we've managed to reconstruct post-incident, the tug was allowed to dive a little through the turn.A large amount of slack developed in the line that we weren't able to see from our vantage point (the camera was off to the side), and we didn't release in time. To put it more properly, I was just opening my mouth to say "I think we should relea..." when the slack came out of the line, the tug stopped flying, and the ASK was jerked in the flattest spin I've ever seen from a non-wingeron model. To Ryan's credit, he had enough time to recover from the spin, and would have pulled out if he'd had 10 feet more altitude...it was that close. Damage to the sailplane was limited to a bent wingrod, some popped servos, and some cracks around the base of the fin, but we were finished for the day. I landed the tug without a problem.
I put another few sessions on the tug over the next week, and found that I couldn't get the belt drive to make that horrible noise unless I tried to hover the tug (not exactly type-realistic, but oh well). Additionally, the belt seemed to be stretching over time, and there were little flecks of belt on the faceplate of the motor mount. I tried to realign the mount as best I could, closed up tension on the belt, and added a bit of goop adhesive between the motor case and the stabilizing setscrews.
At that point disaster struck yet again, though this time it was entirely my fault. After reassembling the drive system, I mistakenly plugged the sensor harness from the motor into the speed control with one pin misaligned. I couldn't see into the fuselage to hook up the speed control, and while the sensor harness plug is polarized to prevent it being plugged in upside down, the pins are evenly spaced. As soon as I hooked up the drive battery to test the motor, the speed control started letting off a cloud of noxious magic smoke. It took me a while to realize what had happened, but I quickly remembered a string of curse words that I had been neglecting for a while. The speed control definitely needed repair, so off it went to Aveox. Luckily I had a Hacker Master 48 Opto controller on hand for another project, so it was pressed into duty with the motor. The sensored Aveox motor worked happily with the sensorless Hacker controller, so the project could at least move on. Subsequent tests of the driveline proved fine, and briefly hovering the model at the field proved noiseless. Things were looking up.
Additionally, Ryan borrowed an AC/DC current clamp from work, and we were able to get some real-life readings as to what the system was drawing. Seeing as how the motor and the batteries were still coming out cool (motor avg 35C, battery avg 65C *no cooling*), we realized that we could still prop for more current. These are summarized in the following table:
|
| Propeller | Current (MotoCalc) | Current (actual) | RPM @ Prop (MotoCalc) | RPM @ Prop (Actual)
| Top Flite Power Point 22"x10 (wood) | 43.1 (way over) | 30.5 A | 4995 (good) | 5000
| Zinger 20x11 (wood) | 34.7 (close) | 31A | 5164 (good) | 5050
Undeterred, Ryan worked hard that week to repair his sailplane, and we were ready to go by the following weekend (and had equally good weather). I added more throw to the control surfaces on the tug, and rebalanced to a more rearward c/g to further aid controllability. After committing the mantra "release, release, release" to permanent memory, we lined the models up for another tow.
Things seemed uneventful on the climbout, but the awful grinding noise had returned. About 100 feet up and 300 feet out, the tug seemed to suddenly lose power. The sailplane was sinking lower and lower behind the tug, so Ryan dumped the tow as quickly as possible. He recovered from a stall only barely in time to scrape back over the field boundary, but I wasn't so lucky. The motor still sounded fine, but I couldn't climb at all. With a 22" windmilling prop and a high-drag airplane, there was no way I was going to make it back to the field. I dumped the tow line, and tried to bring the tug in as level as possible. The subsequent landing in the reeds (our field is a semi-dry lakebed) cleaned the landing gear off the tug, but didn't harm it to any great degree beyond removing the main landing gear bolt plate.
The problem turned out to be the belt drive...the belt had failed. More specifically, the toothed outer layer of the belt had sheared off at the reinforcement layer, where a group of fibreglass bands reinforce the perimeter of the belt. Feeling very defeated at that point, we nonetheless vowed to get to the bottom of these problems.
The tug repairs were simple, requiring only that I install a new bolt plate and remount the tail wheel bracket. To be safe, I cut a piece of 1/4" ply that was about twice as big as the original plate, and epoxied that in place with West Systems resin and a high-strength filler.The bigger problem was the belt drive; with no spare belt, we were truly out of commission until spares arrived. I exchanged a couple of emails with Tom Hunt again, and he suggested that we try some larger diameter pulleys to spread the load over more teeth. Tom kindly offered to swap the pulleys out for free as long as we returned the ones we had, which we were more than happy to do.
Seeing as how the pulleys were going back anyway, we decided to drop the ratio at the same time. Using an even bigger prop wasn't an attractive option given the problems we'd been having, but dropping the ratio would get the current up and up the revs at the motor end. Though the 1817 Aveox motors are technically a 30A continuous motor, it seemed to be happy to take a bit more load than we were asking of it. A tow is usually rather short, and cooling for the motor couldn't get much better than we had it right now. We ultimately decided to swap the original 2:1 ratio pulleys (30 tooth spur, 15 tooth pinion) for ones with a 1.6:1 ratio (40 tooth spur, 25 tooth pinion). MotoCalc predicted a 10A jump in current for the swap, which was about what we were going for.
At the same time as I was doing this, Ryan took the belt drive to work and used a set of calipers on the main shaft of the belt drive. He noticed that the attachment point for the collet adaptor was a little tapered over its length (0.2500" at the base to 0.246" at the tip), which would have accounted for the vibration we were seeing. After Tom confirmed that this taper wasn't intentional, he agreed to ship us a new shaft along with the new belt and pulleys. I anxiously awaited their arrival.
+905521:Working out the Bugs
I got a box in the mail from Tom a couple of weeks later, and couldn't wait to get back at this project (which was proving more challenging than originally had been hoped!). In the mean time, our APC-E props had also arrived, which gave us more test subjects to work with. After reassembling the drive with the parts, I fired the drive up at home. To my immense satisfaction, everything worked flawlessly! The drive was smooth and vibration free with any of the props, and the new ratio was pulling very well. I didn't do much testing there since I didn't have a current clamp, but Ryan promised to bring one with him to the field that weekend.
Stunningly, the weekend brought good weather to Calgary yet again, so I trundled off to the field in high spirits. Thinking that MotoCalc might have messed up the numbers again, I used the smaller 20x11 Zinger prop to do a quick test flight before Ryan showed up. The Telemaster was a real powerhouse now, so I never bothered to use more than half throttle for the whole flight. This probably saved the 48A Hacker speed control, because the current clamp showed some surprising numbers when we tested the motor on the new ratio. For example, the tug was pulling like a demon, but now even the 20x11" was drawing 65A! I'm amazed that the controller held up, but it seemed fine (fortune favors the bold, I guess). Note that the Aveox controller is rated to 60A continuous/100burst and would be fine at these currents, but the little Hacker controller is only rated to 48A! At any rate, I doubt the motor would appreciate 65A, and we were shooting for 40.
Seeing as how we didn't have any smaller props with us, we decided to chop the tips off the Zinger to try to get the current under control. Taking an inch off each blade got the current down to 40A, and the pull was still much better than the 22x10" on the old 2:1 ratio. The new numbers are as follows:
|
| Propeller | Current (MotoCalc) | Current (actual) | RPM @ Prop (MotoCalc) | RPM @ Prop (Actual)
| Zinger 20x11 (wood) | 56.3A (way under) | 65A! | n/a | n/a
| Zinger 18x11 (cut down 20x11) | 41.9A (good) | 40A | 6274 (way over) | 5600
| APC-E 20x13 | 56.4A (WAY over) | 40.5A | 5900 (way over) | 5200
+905522: Flying the Sailplane on Tow
The subsequent flights were all that we'd hoped for when we started the project started, and more. While the first group of flights was done on the 22x10 Power Point, the nicest match we found was the cut down "18x11" Zinger. It seemed to have good static thrust, and the extra bit of pitch speed helped keep the models moving efficiently. The combo now had power to spare, the climb rate was good, and the motor and drive battery come out cool (38C/65C respectively). I still had difficulty turning the tug, but was able to keep Ryan's ASK above me even at solid rates of climb.
A couple of notes about flying the sailplane on tow might be helpful. First off, it's nice to have someone keep the wingtip off the ground for you as the sailplane starts moving. On tarmac or very short grass, this might not be necessary, but it's probably still a good safety precaution. Don't do as the fellow did in video #5 however, and actually hold onto the tip: that can cause problems pretty quickly. Remember to "fly" the ailerons at all speeds to keep the wingtips out of the grass.
On climbout, it's probably going to be necessary for the sailplane to add in a bit of "up" trim as compared to trimmed level flight. Uncouple the ailerons and rudder if you fly that way normally, and mainly work on keeping the wings level with the ailerons. A powerful tug will do much of the work in the turns, but if you find that you get any slack in the turns on tow, it might be helpful to work to keep the sailplane solidly to the outside of the turn with a little bit of rudder. It will really help you maintain a good tow if both pilots communicate with each other the whole way up.
On the other hand, the relative vertical position of the sailplane compared to the tug is more the tug pilot's responsibility. Tows should always be done at full power unless the sailplane is tiny or fragile, and the speed of the tow (and relative height of the sailplane) can be controlled by how quickly the tug climbs. If the sailplane is constantly below the tug, the climb rate should be shallower. If the sailplane threatens to overtake the tug, the climb rate can be steepened.
Finally, if ANY problems occur: "release, release, release"
+905523: Releasing the Line
At the release point, the pilots should communicate the exact moment that the sailplane attempts to let go. That way, if a problem occurs, the tug can drop the line instead. Never attempt to turn either model without having confirmed that a clean release has occurred. A small flag made out of neon surveyors tape was useful for us in recognizing when the sailplane had dropped the line.
If either model has to make a landing approach carrying the line, don't tempt fate by overflying any fences at low altitude. Snagging the line on the ground is a potential disaster...I've seen it happen.
@905515:This Byron Husky suffered extensive damage to its undercarriage and wing struts after catching its towline on a fence while climbing out. The sailplane it was towing released early in the takeoff roll after getting sideways on tow, and the tug pilot forgot that he was still carrying the towline.
Thoughts on Using the Senior Telemaster as a Sailplane Tug
The Senior Telemaster has the advantage of being a simple kit to assemble, but it really isn't an ideal tow platform. As finished here, the model is slow, has a limited speed range, and isn't particularly strong. Additionally, while tug pilots with limited experience flying powered models (ie, shanghaied sailplane pilots) might appreciate the generous inherent stability of the Telemaster when flying the model alone, it does make towing a bit more challenging. In case anyone is interested in trying a similar project, let's look at these issues one at a time:
<EM>"The model is slow...with a limited speed range"</EM>
Frankly, it would be hard to change this much on a Senior Telemaster. The airfoil would have to be dramatically changed, and this makes working from a kit a debatable expense. For the truly dedicated, I'd suggest a good symmetrical section, equipped with flaps and ailerons to improve the potential speed range. I'd think that one of the new Eppler sections developed for giant scale aerobatic sailplanes would be a good choice (ie E166). What's good for the goose should be good for the gander, as they say!
<EM>"Isn't particularly strong"</EM>
The modifications suggested in the article should cure most of these ills, the biggest being that the landing gear bolt plate really needs shoring up with the taller landing gear. Additionally, a plywood plate is a good idea for the tail wheel mount. The thin ply reinforcement of the battery area wasn't wasted weight, but could be carried over to sheeting the floor as well.
One area that might use some shoring up is the wing spar, although this will vary from pilot to pilot. While sport flying the model near the end of our flying season here, I really learned to enjoy flying aerobatics with such a large model. Loops are easy given the ridiculous amount of power available, and the large diameter prop means that there isn't a lot of speed variation around most loops (they're possible from level flight at about 1/4 throttle!). Stall turns and spins are a lot of fun, as the rudder is very effective with the added area. Inverted flight is possible with generous elevator input and quite a bit of throttle, but the model requires a lot of attention in that attitude since it really wants to roll back upright. While rolls are very barrel-like even with generous differential, they have a certain amount of "full-size Cub" appeal that always entertains spectators.
The point of all this discussion was that it's very easy to get carried away - as I did - and overstress the spar. Coming out of the bottom of a relatively tight loop, I heard a definite crack and noticed that I now had a couple of extra degrees of dihedral! I quickly set it down as gently as possible, and noticed that the clear hockey tape that I use to hold the wing halves together had sheared along the entire top surface. More disastrously, one of the top wing spars had failed in compression near the end of the joiner tube. The kevlar spar wrap had probably saved the model, as it was the only thing holding the spars in alignment when I landed. So, for all of you who insist on flying your tug like an overpowered Extra 300 (and who can blame you, really), reinforce that spar!
<EM>"...generous inherent stability..."</EM>
There are a number of ways that the Telemaster could be modified to help it become an easier tug to use.Reducing the wing dihedral to almost none would remove most of the yaw/roll coupling induced by rudder commands, and would allow for easier flight in large-diameter tow circles. Replacing the lifting tail with a flat plate version would remove some of the pitch stability is built into the design as well, but would prevent you from mounting the servos in the stab (cleanly, at least).
The best change that I can see, however,would be to relocate the tow release from behind the wing to over the tug c/g. That would eliminate stabilizing force that the sailplane exerts on the tug, which tends to force it to stay flying in a straight line. This would be difficult on a Telemaster given that the wing must be removable to access the battery, but it should be possible with a little thought.
In the end, however, when does a Telemaster stop being a Telemaster? I'd personally steer prospective tow pilots towards another design, perhaps something along the lines of a Pegasus (as sold by John Derstine at <A href="http://www.scalesoaring.net/">EMM</A>). The Telemaster is an old design, and there are better alternatives available now.
!Conclusion
Looking back on this project now, I can't say that I regret the time and effort that it took. I hope that this has showed how electric tugs are quite feasible even with off-the-shelf components, and a minimum of technical wizardry. Their biggest drawback is still the limited power-on time that is available, but even on inexpensive nicads, two tows are possible with the above setup. With a couple of batteries at the ready, even a single tug should be able to serve the needs of a small sailplane group with minimal delay times.
While the driveline had teething problems, they were more bad luck then a result of a flawed design, and I would have no problem going with the same components again. With the new main shaft, the belt drive is smooth and quiet, and the belt has shown none of the wear that happened with the first belt. The Aveox motor and controller worked very well, and have proven to be very reliable.
The least endearing part of the system has been the Telemaster airframe. Even taking into account some of the findings above, a Telemaster is best suited to towing slower scale models, like the vintage Minimoa below (which has one flying speed - slow). Fans of modern ships will find a cleaner, faster tug to be more useful.
@905516:Author with his <A href="http://www.geocities.com/sunbirdz/Minimoa_Restoration.html">restored</A> 4m Minimoa, originally built from the Chris Williams plan.
@905517:Call me biased, but I think the tapered gull planform of the Minimoa makes it one of the nicest vintage sailplanes. The bubble canopy isn't scale for these markings, and will be replaced in the future.
If anyone is interested, look to a future article (or the forums) for updates on a new tug that I have underway. It will be a purpose-designed electric model, addressing all the problems that I've discussed here (the Telemaster is already sold and serves as an aerial camera platform in its new life). If anyone has any questions or suggestions. I'm happy to entertain them.
Until then!
While the wing is originally designed to be a single-piece unit, I wanted a two-piece wing to make transporting and storing it a lot easier. I ended up using a 5/8" solid aluminum rod as my joiner rod, since I have a lot of it on hand for when I build sailplane wings. The tube that sits in the wing is a 5/8" inner diameter carbon tube, available from Aerospace Composites (ACP) or The Composite Store (CST).
The finished wing is taped together before flight, and held on with the kit-specified rubber bands. I generally prefer a bolt-on wing arrangement, but I thought that bands would make quick battery changes a little easier. Do remember to use enough rubber bands to keep the wing secured.
Since I didn't have any angled rod in stock and the Telemaster wing has a fair amount of dihedral in it, I ended up using a fairly short wing joiner tube. Given that I ended up having problems with wing strength and that the dihedral isn't required for a tug anyway, if I built another Telemaster I would remove a fair amount of dihedral and use a longer joiner tube (maybe three bays per wing instead of only two). The joiner rod itself never bent at all.
I also changed the shear web arrangement from the kit spec, and used vertical grain balsa between the spars.Webs glued to the face of the spar are inefficient, since they really only test shear strength of the glue and the balsa to the depth that the glue penetrates. The first third of the wing used balsa the full width of the spar, and then I gradually tapered the thickness of the web towards the tip, which used only 1/16" balsa webs. Be sure to keep the webs centered in the middle of the spar. Finally, I wrapped the spar from the root to one bay beyond the end of the joiner tube with Kevlar thread, available from ACP or CST. This helps to prevent the spar from bursting at the ends of the tube, where such stresses are typically the greatest.
@905504:Since the joiner rod is straight, the joiner tube must sit in the wing at an angle.
I don't like torque rods of any description, so I decided to mount the servos in the wing (Multiplex MCV2 digitals, which were on hand). Mounting them was a simple affair, consisting of a couple of balsa braces. I decided to leave the servos exposed to the airstream instead of concealed in the wing, since this is a powered model after all.
@905505:Blocking the aileron servo in place.
@905506:A small triangular patch of balsa used to lead out the aileron extension. This ended up being too close to the outside of the fuselage, and so the wire was eventually led out against the root rib. Note the kevlar spar wrapping.
!!Final Assembly
The finished structure is light and efficient when built stock, so I didn't go to many great lengths to lighten the structure. I cut lightening holes in the tail ribs with only a very slight benefit in weight, and gave up after lightening a few wing ribs simply because the balsa was hard enough that the Exacto knife hurt my fingers after a while! (I used them anyway)All of this seems like sacrilege given what we've learned about electric flight, but remember that a heavier tug actually works to our advantage in the air, since a heavy tug tracks better on tow. Also try to remember that saving even 4 ounces on a 200+ oz model is of minimal benefit, especially if it takes hours to cut new ribs or lightening holes.
@905507:The finished model "in the bones", on the stock aluminum gear. Getting to this stage took me about 5 sessions in the shop, as building proceeds quickly.
I chose to cover the model in various colours of Ultracote, which took almost as long to do as building the airplane in the first place (those panel lines were all cut from self-adhesive trim sheet). I like Ultracote for its ease of application, and because it doesn't get as brittle as Monokote as it ages (or in low temperatures).
While it would have been easier to do so earlier, after covering the airplane I sheeted the majority of the batter compartment with 1/32" ply. There really isn't much wood available to support the internal components mounted there, and the fuselage needed a bit of stiffening in that area.
Next up was the pack, which I assembled using six sticks of five cells soldered end to end. The wrappers on GMVIS cells can be tricky to remove, but with patience it can be done cleanly and reasonably quickly (hint: try not to tear the wrapper, or it will shred and leave remnants on the cell that are hard to remove without Goo-Gone). Each stick was then wrapped in 1 mil Kapton tape individually, and the sticks were banded together with another wrap of the tape. This treatment is usually reserved for extreme current applications, but I find that all cells benefit from the increased cooling that Kapton allows, and Kapton is much safer since it never melts (within reason; if the pack gets to 500F, the tape melting isn't your biggest problem).
Mounting the battery is of utmost importance on a big model, since a 30 cell pack weighs just under four pounds and can carry a lot of momentum. Given the cost of a big pack for this airplane, I elected to make life easy by not making the pack removable, and blocked it in place with balsa sticks reinforced with thin ply and balsa gussets. It wouldn't be hard to do so, however. One system would be to use filament tape (or Kapton tape for high-current tugs) to attach angle brackets made of ply or plastic to the battery pack, and to retain the pack by screwing these angle brackets to matching counterparts in the fuselage. That system would also allow easy changes of c/g, if the fuselage track had a series of matching bolt holes.
@905508:The pack mounted in the airplane. Balsa gussets were added to the ply battery hold-downs after I started flying the model aerobatically, mainly as a precaution.
I use 4mm gold pins for my pack connections, since I have them on hand from high-current projects. The Sermos connector pictured below merely adapts the pack to my charger leads.
@905509:The Aveox 1817/4Y motor fitted to one of Tom Hunt's ModelAirTech H1500 belt drives. The pulleys here represent the earlier 2:1 ratio.
@905510:A scoop is usually fitted over the hole in the bottom of the fuselage, but it had been temporarily removed here for access to the speed control wires.Bare_Bones_Stock_Gear.jpg"></A>
Since the Aveox speed control is an opto unit, I used a receiver pack made up from five 1100mah Sanyo AA cells. The receiver is a Multiplex IPD 9 channel model, and I fly with a Multiplex Profi 3030 transmitter.
!Flying
After the tug was finally completed, a couple of run-up tests were performed to verify that the drive system was working properly. Unfortunately, the driveline started to make a horrible vibrating sound at about half throttle, which made me nervous about running it any harder. At this point the belt drive was setup as recommended, with the motor lightly contacting the four stabilizing setscrews and a single zip-tie maintaining the contact. After painstakingly balancing the 22x10 Top Flite Powerpoint prop (no difference), I was getting discouraged. A quick email to Tom Hunt suggested that the prop might not be tracking properly, and since I was able to see that the blades weren't perfectly tracking up, I went back to my local hobby store to find another prop.
@905511:The Telemaster is a good looking design, and I'm especially happy with the way the covering job turned out.
I came home with a Zinger 20x11 wood prop, and again balanced it as best I could. No difference, the vibration was still there. I could feel a bit of play between the output shaft on the belt drive and the collet-style prop adaptor, so I asked Tom to send me a new adaptor (which he did without hesitation). Tom mentioned that he had never had any problems with these drives, and that they should operate smoothly throughout the whole RPM range as long as the motor spun at less than 20,000 revs (we were only spinning 10,000), so I had hope that we would be able to lick the vibration issue. When the new adaptor arrived, the vibration was a little lessened, but still not gone. Thinking that a wrap of tape around the output shaft (under the collet) might help, I added a single layer of Kapton tape under the adaptor, and that seemed to get rid of most of the problem. I also ordered some of the new 20x13" APC-E props, thinking that a molded plastic prop would be more likely to run true than a wooden version (which might have warped due to the low humidity that Calgary is famous for).
@905512:The view from behind shows how tall the gear is. The model didn't roll very well through rough grass on the original 3" Robart tires shown here, and they were later changed to 4.5" Sullivan Featherlites.
At that point, I decided to get another set of eyes on the vibration issue, and so invited Ryan over to give his opinion. After much testing and debate, Ryan noticed that the motor still seemed to be vibrating against the stabilizing setscrews. On his suggestion, we added two zip ties in place of the original single one, and another pair around the end of the motor (for four in total). This allowed us to run the motor up to full power with either prop, and the model looked to be ready for a test flight.
The first flights took place at a local pylon racing field, which had a nice concrete runway that we could use. Thanks again to the Calgary<A href="http://www.cmpra.com/"> Miniature Pylon Racing Association </A>for use of their field! The wind was blowing at about 25kph gusting to 35 or so, which was not exactly calm, but I was confident that the 13 lb model could handle the wind. Taxiing was impossible on the high gear and small wheels, but Ryan kindly offered to hold the model in place until I was ready for takeoff.After slowly advancing the throttle to half, the Telemaster was off the ground in about 5 feet! At that point it was climbing strongly, but not making much progress forward. I could tell that the model was going to be a slow flier on the 22x10 prop, but hoped that it would okay in less wind.
The rest of the 4 min flight was largely uneventful, since the wind didn't allow for much beyond a couple of circuits on the field. Landing was easy given the low groundspeed, but I held a bit more speed than normal into the flair to prevent the model from ground looping. Even with that precaution, I was thankful that Ryan was there to catch the wingtip at the end of the flight, since the Telemaster tried to blow over after coming to a stop! That high gear is nice for prop clearance, but can make the airplane a bit tippy on the group.
@905513:The Telemaster coming in on final at the CMPRA field.
@905514:I really liked the look of the model in the air, and it seemed stable even in very active air.
Since the tug seemed to be ready to go, it was time to see how the model behaved with a sailplane behind it. The next weekend was warm and calm (unusual for Calgary in the fall), and seemed like the perfect time to try towing. We were originally going to try towing from the local full-size sailplane field, but it was being used on that weekend for normal operations. The backup location was our own model club field, which was a little rough, but serviceable. The grass was reasonably short since it was late in the year, and while the ground was a little rough (it isn't intended to service wheeled models), the big airplanes seemed to handle it okay.
As Ryan readied his ASK for flight, I took the chance to put a couple of more orientation flights on the tug. It was obviously not lacking in power, coming off the ground in 10 feet at 1/3 throttle (!), but the controllability also seemed good for such a large model as well. There wasn't a noticeable change in pitch trim when the throttle was advanced, and the balance point seemed to be only slightly forward on the C/G. Things seemed to be good to go, so I borrowed Ryan's Schulze 630d charger to top up the pack, and went to work in putting together some tow lines.
We decided to use 200lb braided Dacron tow lines for a couple of reasons, not the least of which was that it was the same line that I use on my winch. In addition to being available, the Dacron line offered a nice balance between strength, stretchiness, and visibility. 80lb monofilament fishing line loops were used on either end of the line to allow it to be fed into the tow releases. It would be a good idea to have a swiveled fishing snap on the sailplane end of the line, since that would allow for quick attachments to loops that are pre-installed in the sailplane's release. Additionally, when the loops are released under load, the cam-style tow releases tend to nick the monofilament loops, which means that replacing them every flight isn't a bad idea. 120-150 foot lines are the norm at our field.
+905518:First Steps
After finding the patch of ground that was most level and free from badger holes, we hooked the models together and pointed them into the wind (2-5mph). The plan was to release after making one large of the field, with the first turn to the left. It's always better for the tug and sailplane pilots to agree upon a flight plan before taking off,especially when both pilots are relatively inexperienced.Additionally, be sure to release at the first sign of trouble, since things can get ugly quickly.
After taking a couple of deep breaths, it was time to go! Smoothly but quickly advancing the throttle, the combo accelerated strongly. There was no problem keeping either the sailplane or the tug tracking straight, and the climb rate seemed to be good. Unfortunately, the belt drive started making ugly sounds again (see Video 1), though they were only under the load of a sailplane.
After taking off, we proceeded to climb out in a straight line to the southwest. We had always been told to tow at full throttle, and to modulate the speed of the tow by climb angle, but this was proving difficult in practice. I was having trouble keeping the tail of the tug up, and so Ryan decided to release before things got too slow. We had managed to get to about 300 feet quite quickly, but since the sailplane was well below the tug, it seemed safer to try again after making some adjustments.
The sailplane had enough time to make a couple of passes over the lakebed, and even wandered though a promising patch of lift. However, since it was only his second flight with the airplane, Ryan wisely decided to land instead. After making a low pass to drop the towline, I dropped the tug in for a gently landing soon thereafter. While I could have elected to land with the line still attached, I feel that it's safer to drop it instead. After watching a tug snag a fenceline with its towline (and subsequently pancake into the deck really hard), I've always followed this principle.
+905519: Trying Again
After recharging the battery (it took about 500mah, so this wasn't strictly necessary), we lined the models up for another tow. I increased the amount of down elevator travel that I had available, and added some down trim to the tug as well. The second takeoff went as easily as the first, though in a more coordinated fashion. Since I have trouble watching two models at once, I had asked Ryan to tell me when his sailplane was off the ground. This prevents the dangerous situation where the tug takes off first, before the sailplane is fully up to speed. A well-behaved sailplane like the ASK-21 might not object to be dragged around barely above stall speed, but an aerobatic model like aFox surely won't appreciate it. At any rate, I "flew" the tug along the ground until I heard the call from Ryan, and there were no problems at all.
On climbout, I still found that I had to hold some down elevator, but the climbout was much more controllable in pitch than it had been before. A new problem arose with the first turn however...the tug didn't want to change course! I needed full aileron and a good deal of rudder to coordinate the turn, which I attributed both to fighting the Telemaster's inherent stability and the right moment from the tow release (this would be eliminated by towing from on the tug c/g instead of from behind it). Consequently, I inadvertently let the tug dive a little bit through the turn as I concentrated on keeping the turn large and...
+905520: Things Get Ugly
Ugly is putting it mildly. From what we've managed to reconstruct post-incident, the tug was allowed to dive a little through the turn.A large amount of slack developed in the line that we weren't able to see from our vantage point (the camera was off to the side), and we didn't release in time. To put it more properly, I was just opening my mouth to say "I think we should relea..." when the slack came out of the line, the tug stopped flying, and the ASK was jerked in the flattest spin I've ever seen from a non-wingeron model. To Ryan's credit, he had enough time to recover from the spin, and would have pulled out if he'd had 10 feet more altitude...it was that close. Damage to the sailplane was limited to a bent wingrod, some popped servos, and some cracks around the base of the fin, but we were finished for the day. I landed the tug without a problem.
I put another few sessions on the tug over the next week, and found that I couldn't get the belt drive to make that horrible noise unless I tried to hover the tug (not exactly type-realistic, but oh well). Additionally, the belt seemed to be stretching over time, and there were little flecks of belt on the faceplate of the motor mount. I tried to realign the mount as best I could, closed up tension on the belt, and added a bit of goop adhesive between the motor case and the stabilizing setscrews.
At that point disaster struck yet again, though this time it was entirely my fault. After reassembling the drive system, I mistakenly plugged the sensor harness from the motor into the speed control with one pin misaligned. I couldn't see into the fuselage to hook up the speed control, and while the sensor harness plug is polarized to prevent it being plugged in upside down, the pins are evenly spaced. As soon as I hooked up the drive battery to test the motor, the speed control started letting off a cloud of noxious magic smoke. It took me a while to realize what had happened, but I quickly remembered a string of curse words that I had been neglecting for a while. The speed control definitely needed repair, so off it went to Aveox. Luckily I had a Hacker Master 48 Opto controller on hand for another project, so it was pressed into duty with the motor. The sensored Aveox motor worked happily with the sensorless Hacker controller, so the project could at least move on. Subsequent tests of the driveline proved fine, and briefly hovering the model at the field proved noiseless. Things were looking up.
Additionally, Ryan borrowed an AC/DC current clamp from work, and we were able to get some real-life readings as to what the system was drawing. Seeing as how the motor and the batteries were still coming out cool (motor avg 35C, battery avg 65C *no cooling*), we realized that we could still prop for more current. These are summarized in the following table:
|
| Propeller | Current (MotoCalc) | Current (actual) | RPM @ Prop (MotoCalc) | RPM @ Prop (Actual)
| Top Flite Power Point 22"x10 (wood) | 43.1 (way over) | 30.5 A | 4995 (good) | 5000
| Zinger 20x11 (wood) | 34.7 (close) | 31A | 5164 (good) | 5050
Undeterred, Ryan worked hard that week to repair his sailplane, and we were ready to go by the following weekend (and had equally good weather). I added more throw to the control surfaces on the tug, and rebalanced to a more rearward c/g to further aid controllability. After committing the mantra "release, release, release" to permanent memory, we lined the models up for another tow.
Things seemed uneventful on the climbout, but the awful grinding noise had returned. About 100 feet up and 300 feet out, the tug seemed to suddenly lose power. The sailplane was sinking lower and lower behind the tug, so Ryan dumped the tow as quickly as possible. He recovered from a stall only barely in time to scrape back over the field boundary, but I wasn't so lucky. The motor still sounded fine, but I couldn't climb at all. With a 22" windmilling prop and a high-drag airplane, there was no way I was going to make it back to the field. I dumped the tow line, and tried to bring the tug in as level as possible. The subsequent landing in the reeds (our field is a semi-dry lakebed) cleaned the landing gear off the tug, but didn't harm it to any great degree beyond removing the main landing gear bolt plate.
The problem turned out to be the belt drive...the belt had failed. More specifically, the toothed outer layer of the belt had sheared off at the reinforcement layer, where a group of fibreglass bands reinforce the perimeter of the belt. Feeling very defeated at that point, we nonetheless vowed to get to the bottom of these problems.
The tug repairs were simple, requiring only that I install a new bolt plate and remount the tail wheel bracket. To be safe, I cut a piece of 1/4" ply that was about twice as big as the original plate, and epoxied that in place with West Systems resin and a high-strength filler.The bigger problem was the belt drive; with no spare belt, we were truly out of commission until spares arrived. I exchanged a couple of emails with Tom Hunt again, and he suggested that we try some larger diameter pulleys to spread the load over more teeth. Tom kindly offered to swap the pulleys out for free as long as we returned the ones we had, which we were more than happy to do.
Seeing as how the pulleys were going back anyway, we decided to drop the ratio at the same time. Using an even bigger prop wasn't an attractive option given the problems we'd been having, but dropping the ratio would get the current up and up the revs at the motor end. Though the 1817 Aveox motors are technically a 30A continuous motor, it seemed to be happy to take a bit more load than we were asking of it. A tow is usually rather short, and cooling for the motor couldn't get much better than we had it right now. We ultimately decided to swap the original 2:1 ratio pulleys (30 tooth spur, 15 tooth pinion) for ones with a 1.6:1 ratio (40 tooth spur, 25 tooth pinion). MotoCalc predicted a 10A jump in current for the swap, which was about what we were going for.
At the same time as I was doing this, Ryan took the belt drive to work and used a set of calipers on the main shaft of the belt drive. He noticed that the attachment point for the collet adaptor was a little tapered over its length (0.2500" at the base to 0.246" at the tip), which would have accounted for the vibration we were seeing. After Tom confirmed that this taper wasn't intentional, he agreed to ship us a new shaft along with the new belt and pulleys. I anxiously awaited their arrival.
+905521:Working out the Bugs
I got a box in the mail from Tom a couple of weeks later, and couldn't wait to get back at this project (which was proving more challenging than originally had been hoped!). In the mean time, our APC-E props had also arrived, which gave us more test subjects to work with. After reassembling the drive with the parts, I fired the drive up at home. To my immense satisfaction, everything worked flawlessly! The drive was smooth and vibration free with any of the props, and the new ratio was pulling very well. I didn't do much testing there since I didn't have a current clamp, but Ryan promised to bring one with him to the field that weekend.
Stunningly, the weekend brought good weather to Calgary yet again, so I trundled off to the field in high spirits. Thinking that MotoCalc might have messed up the numbers again, I used the smaller 20x11 Zinger prop to do a quick test flight before Ryan showed up. The Telemaster was a real powerhouse now, so I never bothered to use more than half throttle for the whole flight. This probably saved the 48A Hacker speed control, because the current clamp showed some surprising numbers when we tested the motor on the new ratio. For example, the tug was pulling like a demon, but now even the 20x11" was drawing 65A! I'm amazed that the controller held up, but it seemed fine (fortune favors the bold, I guess). Note that the Aveox controller is rated to 60A continuous/100burst and would be fine at these currents, but the little Hacker controller is only rated to 48A! At any rate, I doubt the motor would appreciate 65A, and we were shooting for 40.
Seeing as how we didn't have any smaller props with us, we decided to chop the tips off the Zinger to try to get the current under control. Taking an inch off each blade got the current down to 40A, and the pull was still much better than the 22x10" on the old 2:1 ratio. The new numbers are as follows:
|
| Propeller | Current (MotoCalc) | Current (actual) | RPM @ Prop (MotoCalc) | RPM @ Prop (Actual)
| Zinger 20x11 (wood) | 56.3A (way under) | 65A! | n/a | n/a
| Zinger 18x11 (cut down 20x11) | 41.9A (good) | 40A | 6274 (way over) | 5600
| APC-E 20x13 | 56.4A (WAY over) | 40.5A | 5900 (way over) | 5200
+905522: Flying the Sailplane on Tow
The subsequent flights were all that we'd hoped for when we started the project started, and more. While the first group of flights was done on the 22x10 Power Point, the nicest match we found was the cut down "18x11" Zinger. It seemed to have good static thrust, and the extra bit of pitch speed helped keep the models moving efficiently. The combo now had power to spare, the climb rate was good, and the motor and drive battery come out cool (38C/65C respectively). I still had difficulty turning the tug, but was able to keep Ryan's ASK above me even at solid rates of climb.
A couple of notes about flying the sailplane on tow might be helpful. First off, it's nice to have someone keep the wingtip off the ground for you as the sailplane starts moving. On tarmac or very short grass, this might not be necessary, but it's probably still a good safety precaution. Don't do as the fellow did in video #5 however, and actually hold onto the tip: that can cause problems pretty quickly. Remember to "fly" the ailerons at all speeds to keep the wingtips out of the grass.
On climbout, it's probably going to be necessary for the sailplane to add in a bit of "up" trim as compared to trimmed level flight. Uncouple the ailerons and rudder if you fly that way normally, and mainly work on keeping the wings level with the ailerons. A powerful tug will do much of the work in the turns, but if you find that you get any slack in the turns on tow, it might be helpful to work to keep the sailplane solidly to the outside of the turn with a little bit of rudder. It will really help you maintain a good tow if both pilots communicate with each other the whole way up.
On the other hand, the relative vertical position of the sailplane compared to the tug is more the tug pilot's responsibility. Tows should always be done at full power unless the sailplane is tiny or fragile, and the speed of the tow (and relative height of the sailplane) can be controlled by how quickly the tug climbs. If the sailplane is constantly below the tug, the climb rate should be shallower. If the sailplane threatens to overtake the tug, the climb rate can be steepened.
Finally, if ANY problems occur: "release, release, release"
+905523: Releasing the Line
At the release point, the pilots should communicate the exact moment that the sailplane attempts to let go. That way, if a problem occurs, the tug can drop the line instead. Never attempt to turn either model without having confirmed that a clean release has occurred. A small flag made out of neon surveyors tape was useful for us in recognizing when the sailplane had dropped the line.
If either model has to make a landing approach carrying the line, don't tempt fate by overflying any fences at low altitude. Snagging the line on the ground is a potential disaster...I've seen it happen.
@905515:This Byron Husky suffered extensive damage to its undercarriage and wing struts after catching its towline on a fence while climbing out. The sailplane it was towing released early in the takeoff roll after getting sideways on tow, and the tug pilot forgot that he was still carrying the towline.
Thoughts on Using the Senior Telemaster as a Sailplane Tug
The Senior Telemaster has the advantage of being a simple kit to assemble, but it really isn't an ideal tow platform. As finished here, the model is slow, has a limited speed range, and isn't particularly strong. Additionally, while tug pilots with limited experience flying powered models (ie, shanghaied sailplane pilots) might appreciate the generous inherent stability of the Telemaster when flying the model alone, it does make towing a bit more challenging. In case anyone is interested in trying a similar project, let's look at these issues one at a time:
<EM>"The model is slow...with a limited speed range"</EM>
Frankly, it would be hard to change this much on a Senior Telemaster. The airfoil would have to be dramatically changed, and this makes working from a kit a debatable expense. For the truly dedicated, I'd suggest a good symmetrical section, equipped with flaps and ailerons to improve the potential speed range. I'd think that one of the new Eppler sections developed for giant scale aerobatic sailplanes would be a good choice (ie E166). What's good for the goose should be good for the gander, as they say!
<EM>"Isn't particularly strong"</EM>
The modifications suggested in the article should cure most of these ills, the biggest being that the landing gear bolt plate really needs shoring up with the taller landing gear. Additionally, a plywood plate is a good idea for the tail wheel mount. The thin ply reinforcement of the battery area wasn't wasted weight, but could be carried over to sheeting the floor as well.
One area that might use some shoring up is the wing spar, although this will vary from pilot to pilot. While sport flying the model near the end of our flying season here, I really learned to enjoy flying aerobatics with such a large model. Loops are easy given the ridiculous amount of power available, and the large diameter prop means that there isn't a lot of speed variation around most loops (they're possible from level flight at about 1/4 throttle!). Stall turns and spins are a lot of fun, as the rudder is very effective with the added area. Inverted flight is possible with generous elevator input and quite a bit of throttle, but the model requires a lot of attention in that attitude since it really wants to roll back upright. While rolls are very barrel-like even with generous differential, they have a certain amount of "full-size Cub" appeal that always entertains spectators.
The point of all this discussion was that it's very easy to get carried away - as I did - and overstress the spar. Coming out of the bottom of a relatively tight loop, I heard a definite crack and noticed that I now had a couple of extra degrees of dihedral! I quickly set it down as gently as possible, and noticed that the clear hockey tape that I use to hold the wing halves together had sheared along the entire top surface. More disastrously, one of the top wing spars had failed in compression near the end of the joiner tube. The kevlar spar wrap had probably saved the model, as it was the only thing holding the spars in alignment when I landed. So, for all of you who insist on flying your tug like an overpowered Extra 300 (and who can blame you, really), reinforce that spar!
<EM>"...generous inherent stability..."</EM>
There are a number of ways that the Telemaster could be modified to help it become an easier tug to use.Reducing the wing dihedral to almost none would remove most of the yaw/roll coupling induced by rudder commands, and would allow for easier flight in large-diameter tow circles. Replacing the lifting tail with a flat plate version would remove some of the pitch stability is built into the design as well, but would prevent you from mounting the servos in the stab (cleanly, at least).
The best change that I can see, however,would be to relocate the tow release from behind the wing to over the tug c/g. That would eliminate stabilizing force that the sailplane exerts on the tug, which tends to force it to stay flying in a straight line. This would be difficult on a Telemaster given that the wing must be removable to access the battery, but it should be possible with a little thought.
In the end, however, when does a Telemaster stop being a Telemaster? I'd personally steer prospective tow pilots towards another design, perhaps something along the lines of a Pegasus (as sold by John Derstine at <A href="http://www.scalesoaring.net/">EMM</A>). The Telemaster is an old design, and there are better alternatives available now.
!Conclusion
Looking back on this project now, I can't say that I regret the time and effort that it took. I hope that this has showed how electric tugs are quite feasible even with off-the-shelf components, and a minimum of technical wizardry. Their biggest drawback is still the limited power-on time that is available, but even on inexpensive nicads, two tows are possible with the above setup. With a couple of batteries at the ready, even a single tug should be able to serve the needs of a small sailplane group with minimal delay times.
While the driveline had teething problems, they were more bad luck then a result of a flawed design, and I would have no problem going with the same components again. With the new main shaft, the belt drive is smooth and quiet, and the belt has shown none of the wear that happened with the first belt. The Aveox motor and controller worked very well, and have proven to be very reliable.
The least endearing part of the system has been the Telemaster airframe. Even taking into account some of the findings above, a Telemaster is best suited to towing slower scale models, like the vintage Minimoa below (which has one flying speed - slow). Fans of modern ships will find a cleaner, faster tug to be more useful.
@905516:Author with his <A href="http://www.geocities.com/sunbirdz/Minimoa_Restoration.html">restored</A> 4m Minimoa, originally built from the Chris Williams plan.
@905517:Call me biased, but I think the tapered gull planform of the Minimoa makes it one of the nicest vintage sailplanes. The bubble canopy isn't scale for these markings, and will be replaced in the future.
If anyone is interested, look to a future article (or the forums) for updates on a new tug that I have underway. It will be a purpose-designed electric model, addressing all the problems that I've discussed here (the Telemaster is already sold and serves as an aerial camera platform in its new life). If anyone has any questions or suggestions. I'm happy to entertain them.
Until then!