HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Oct 07, 2012, 05:11 AM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Discussion
Alternative tail drive 450 raised tail, finally successful

Hi all,

Just a little project I started last week: to eliminate the belt drive from my Heli Artist 450 size BO 105
I had multiple reasons for this:
1)Belt drive is consuming a lot of power due to high friction losses in the extra guide rollers (which in themselves present a possibility for faillure), the air transport occuring in the tail boom at higher belt speeds and the general power loss due continuous bending and stretching of the belt while it runs over the pulleys.
2) tail control feels mushy due to the bowdencable control, and tail control needs painstaking readjustment too, every time the mechanics is removed from the fuselage. Due to the big tail boom and connecting piece, no space for a bellcrank and straight control rods in the tail.
3) belt tension is critical, needs accurate and timeconsuming readjustment every time the mechanics is placed back in the fuselage after maintenance. Not to mention, it is a P in the A to get the belt back over the pulleys with the mechanics in the fuselage.
4) the tailboom is essentially unnecessary weight as the HeliArtist fuselage has plenty structural strength to handle the torque. I would like to get rid of that too.

First I needed to convert the clone T-rex 450 to semi torque tube.
For this I needed to replace the intermediate shaft with pulley for one with a bevel gear. Unfortunately, for that the upper bearing needs to be relocated upwards (by exactly 12 mm) and that was interfering with the position of the elevator servo. In the new position I could can only use the aft screws because the forward screws of the upper bearing block will in the opening of the servo mounting, but that should not present any problem, the bearing is not carrying a big load.
Since the boomblock became redundant, I placed the elevator servo where the boomblock used to be, at the right side of the frame, and established elevator control over a small, ballraced bellcrank that came from a discarded Protech Zoom 450 IC. For this I needed to cut out the opening in the frame, and some material needed t be removed from the swash guide in order to let the controlrod pass. The bellcrank can be fitted in the forward hole of the original servo mounting.

Once that was done, I asked a friend with a milling machine to make me a 22 mm square block with bearing seats, to accomodate the horizontal bevel gear, and placed that between the frame sides. It took some grinding to clear the upper bearing block and the mounting screws of the elevator servo, but nothing too problematic. (In hindsight: the exact dimensions of this block could be reduced to 22 x 15 x 12 mm, with the bearing recesses centered in the 22 x 15 rectangle. Bearing seat dimensions 10 x 3 mm, on both sides of the block)

Next step was to convert the tail rotor to bevel gear drive.
For this I needed the friend with the milling gear again. Essentially, just a small block 19 x 15.5 x 12 mm with bearing seats at the appropriate location did the job The bearing seats should be centered in the 19 x 15mm.5 rectangle, offset by 0.75 mm (measuring 7 mm to one side, 8,5 mm to the other). Same bearing dimensions as above. I hope the photo's explain for themselves....
It is a bit crude, but it functions.
In my case I could not use the braces around the bearings, due to the diameter of the bevel gear, but the bevel gears can be reduced by 1 or 2 mm without any problems. You need to make new braces yourself though. Just 3 mm bushings with a length of 15.5 mm and a threaded bore M2. Should not be a problem for anybody halfway skilled on a lathe....

To get the mechanics drive the tail, I made a "flex-drive" consisting of an aluminium tube, lined with a snug fitting Teflon Tube, in which a 1.2 mm (EDIT: see post #4, testing shows this should be 1.0 mm) spring steel shaft is running. I chose this construction, because back in the old days this solution was used in the old Schlüter Cobra and Graupner Bell 212 Twin Jet, and it seemed to function at twice the RPM and more than 4 times the transported power, so I figured that should work here as well....
(EDIT: quick calculation shows that the torque that needs to be transferred is in roughly in the area of 5 Ncm, which really should not be a problem.)
Also my Vario Bell 47GII has similar construction (not with a radius though) and that tail runs extremely smooth and with less friction losses than a ballraced driveshaft.
In order to reduce friction (and heat generation) as much as possible, I polished the shaft first with a good metal polishing compound, after that with a car wax&polish, and finally I lubricated it with a teflon based lubricant.

For connecting the bevel gears to the flex-drive, two simple adapters were made on the lathe, basically an aluminium bushing of 8 mm diameter with two threaded holes for grubscrews and a part filed flat to 4.5 x 1.5 mm. Careful: Use flat nosed M3 x 3 grubscrews only, preferrable for allen key. Not the pointy ones, and not the ones with the cutting tip either.

Next it was the tedious job of fabricating, aligning and glueing in the frames that would fixate the tube and the tail rotor, as well as the bellcrank for the tail rotor control. I made those out of a good quality 2.5 mm plywood.

In the mean time I made a test rig and tested the flex drive, with a Dremel as driving motor. Tests (up to 10000 RPM) went well enough to continue the modification.

After fitting everything together, I made a few testruns on the ground.
During these test runs I noted that I had to reduce ESC opening from 65 to 55% (Align ESC in governor mode) in order to get the same RPM. This indicates that the first goal, reducing energy loss, has been achieved.

After this, it was time for the first carefull hops to check if swash control was OK (it was) and to check and adjust tail control.
I found that I needed to shorten the tail servo horn considerably, because with only the bowdencable replaced for the rod and bellcrank, I suddenly had almost twice the tail response because no more slop in the control. Tail control was incredibly crisp compared to the bowden cable, almost felt like my KDS 450 SV that has heading lock.
So, second goal achieved as well.

During all this setting up, I needed to remove the mechanics several times, and since removing the mechanics now is a matter of loosening two screws and sliding it forward, that is now an easy job too. third goal achieved as well.

Checking the weight showed a weight reduction of approx 30 grammes, which in itself is responsible for a reduction in current drain of approx 250 mAh. Not much, but still, fourth goal achieved....

So next step: extended hoovering tests....
Unfortunately, after approx 2 minutes hoovering, the shaft snapped,
fortunately I was able to land without additional damage, but this was a major bummer....

The radius at which the flex drive is curving upwards should not present problems to a good quality spring steel, and most likely the wire I used was of the wrong quality, or it had a material defect.
So it is back to the LHS to get a good quality spring steel, as I have seen these shafts work fairly reliable with much bigger load and over 3 times the lenght (yes... I flew that Graupner Bell twin jet myself a long time ago); it should work... Will report back on that as soon as new shaft is installed (will be after the weekend.... )

I hope, the pictures are clear.

Brgds, Bert
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 12, 2012 at 12:21 PM. Reason: additional calculation about torque, some measurements
Reply With Quote
Sign up now
to remove ads between posts
Old Oct 07, 2012, 10:56 AM
Love my scale Whirlybirds
Keyrigger's Avatar
Mississauga, Ont., Can.
Joined Sep 2009
2,380 Posts
That is some excellent work to get away from the belt system. Since there was a failure of the shaft, is there a way that you can alter the path so that the guide tube first drops near the bottom of the boom, then swings up just missing the junction of the upper tail and the boom, leading to achieving a larger radius for the wire to go around? I have seen, in the construction drawings for the Century Huey and Seaking, where they first come from the tail drive output and then bend down before coming up to the tail. You could easily find a market for your system once you get the minor bugs worked up. Well done.

Don
Keyrigger is online now Find More Posts by Keyrigger
Reply With Quote
Old Oct 07, 2012, 01:56 PM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Thanks for the compliment, but the credit goes to old mr Schlüter, who was the originator of the idea (to my knowledge)

I did exactly what you suggested when I bent the prototype tube.... unfortunately, when aligning things during build, I ended up with this straight line, but the radius is not really that much shorter. Radius at the moment is appr 4~5 inch (difficult to measure once it is in the tail) and I could at max gain another half inch, that is the clearance between tube and bottom of the tail.

I am not a metallurgist, but my friend (the guy that made the green parts) is a metal worker. He thinks I did not use a real quality spring-steel, after doing a file-test on the snapped shaft.

It is difficult to market, as I would need to find a way to make templates for all existing tails to shape the tube, make frames, etc etc, only to find that people can do what I did for a few bucks and just instead of modifying the heli, buy a cheap Pro-clone that allready has the Torque Tube drives....

But I have thought about it, and prefer to just share the idea anyway. Because when a hobby becomes a job.... it stops being a hobby usually.
I truly hope, I give some people a nice idea to either utilize their Torque Tube heli in a raised tail, or just modify a belt drive like I did. And maybe, somebody comes up with a better Idea, which I woud gladly use myself.... It is all about getting people to think outside the box, isn't it?

It was a nice few days of tinkering:
Time spent so far:
machining the blocks and connectors: approx 4 hours on Lathe and Mill
Fabricating frames and placing them: approx 2 hrs woodwork and glueing, and 25 hrs waiting for the glue to set (but that time was spent mostly sleeping )
Time spent looking in my old parts heaps and scrap boxes: 15 hrs before I found the bellcranks and other small parts (my hobbyshack is an undiluted chaos)
Money spent: 20 Euro for the gears, 5 Euro for the Alu-tube, Teflon lining and steel wire.

Brgds, Bert

EDIT: having thought about the marketing stuff: that is not really going to be feasible, but if there is interest, I can try to make the drawings for the bearing blocks that I needed for the conversion, for the wire-connectors, and a general arrangement on how to install the tube.
Just let me know, and with suffcient people to make it worth my while (not meant in a commercial way, I will just make some drawings and put them here for everybody to download), together with a description on the points of attention.
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 08, 2012 at 03:13 AM.
Reply With Quote
Old Oct 08, 2012, 04:11 AM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Update: my LHS did not have the 1.2 mm steel wire I started out with, so I opted for 1 mm.
Due to the smaller diameter this should have less stress in the outer fibre and is still capable of transferring the required torque. It does twist a bit more (under static torque of 5 Ncm approx 90 degrees) but the first battery charge has gone through the motor, and power consumption has gone down seriously: Normally a 7 minute hoover would require over 1900 mAh to be charged back, but this time the counter stopped at 1623 mAh..... WOW!
Put a notepad in the heli to keep track of service life of the 1 mm shaft
Only need to make new couplings with a 1 mm bore, these seem to wobble a bit....

Weight reduction is noticeable as well: I had to shift the battery backwards approx 1 inch (it is now touching the mechanics and cant be moved further) but still the helicopter is slightly nose heavy. In standard configuration, with tailboom, belt, and tailservo fitted under the boom, the battery needs to be touching the front window and still it would be slightly tail heavy (and that's a 2600 mAh battery )
Second hoover with timer at 8 minutes gave a recharged capacity of 1990 mAh.
All in all I would say this is promising: Going back from 2600 mAh to 2200 mAh and still get over 7 minutes of flight time becomes an option now.

Brgds, Bert

Edit: darn! 2 minutes into the 3rd battery, the 1.0 mm wire snapped again....
This is getting annoying.... But not giving up, yet.... The shaft snapped at exactly the same location (appr 2 inch from the tail gearbox), and that is not the place of the smallest radius or something like that, so it probably has to do with vibration, unwanted movement or misalignment, rather than radius or torque.
Something to look into deeper.
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 09, 2012 at 02:07 AM. Reason: Update + spelling
Reply With Quote
Old Oct 09, 2012, 07:12 AM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Three shafts further in the process, And thrying to theoriticize the reason for breaking every time at the same point, approximately 1 inch from the end of the upward curve, I start to get the impression that the cause of breaking is axial movement, originating somewhere from the tailrotor: it is exactly where an axial movement would cause a continuous bending and stretching of the shaft

Either unbalance in the tailrotor, or the combination of slight misalignment and slight wobble of the connector.

R&D is continueing, with several solutions in mind.

Brgds, Bert
Brutus1967 is offline Find More Posts by Brutus1967
Reply With Quote
Old Oct 09, 2012, 11:49 AM
Registered User
USA, CA, Red Bluff
Joined Dec 2007
952 Posts
Quote:
Originally Posted by Brutus1967 View Post
Three shafts further in the process, And thrying to theoriticize the reason for breaking every time at the same point, approximately 1 inch from the end of the upward curve, I start to get the impression that the cause of breaking is axial movement, originating somewhere from the tailrotor: it is exactly where an axial movement would cause a continuous bending and stretching of the shaft

Either unbalance in the tailrotor, or the combination of slight misalignment and slight wobble of the connector.

R&D is continueing, with several solutions in mind.

Brgds, Bert
Bert: Have you thought about a cable such as used in a tacometer or speedometer? Just trying to help Ed
prop bender is offline Find More Posts by prop bender
Reply With Quote
Old Oct 09, 2012, 12:25 PM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Yes, have thought about that....

As far as i know, HeliArtist uses something like that in the 600 size BO 105,
but found it not easy to find one in really small diameter (smallest was 2.5 mm and by that size they get heavy) and second problem was how to make adapters for a cable like that.

I am also worried about their ability to withstand high RPM, they are typically designed for max 1000 RPM and I need close to 10.000 RPM.

Finally I am slightly worried that those cables, needing grease for lubrication, do have a rather high power loss due friction.

But it is very well possible, I need to look into that as a last resort.

For now, the plan is to fit a bearing seat at the aft end of the guide-tube, and make the connector ballraced, so the shaft cannot move up, down or wobble.
After some thinking, I realized, that in the Graupner/Helmut Berhhardt designs of the mid 70's, the shaft was clamped into a ballraced connector, opposite to my solution of a free running connector fitting into a slot in the bevelgear wich creates the possibility of "rattling about".

In a few minutes I am off to my friend the metal worker to fabricate some stuff.

To be honest, I did 12 testhoovers of 8 minutes now in just one day, and that endless hoovering is getting boring....

But thanks for thinking along, and if all else fails....

Brgds, Bert
Brutus1967 is offline Find More Posts by Brutus1967
Reply With Quote
Old Oct 09, 2012, 03:01 PM
Registered User
USA, CA, Red Bluff
Joined Dec 2007
952 Posts
I hope you get it sorted out. I really like your idia of getting rid of the belt in the raised tail 450. I fly sport scale 450's, and I agree, the belt is a power robber. I'll be watching your thread closely. wishing you seccess. Ed
prop bender is offline Find More Posts by prop bender
Reply With Quote
Old Oct 09, 2012, 04:11 PM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Will update soon.
Just spent a full evening behind the lathe, made my first real bearing fittings ever (I mean, with a tolerance of 0.01 mm or better).
Time consuming, but worthwile...

What I did was make a small extension on the tube, containing a bearing seat.
Made a new connector, this time with a small elongation, also with bearing fit.

Now the intention is, that the connector driving the tailrotor, is rotating in its own separate bearings, making all unwanted movement of the shaft impossible.
At least, that is is the theory, and if successfull, the shaft should stay in one piece....Weight penalty is only 2 grammes, so not too bad....


Wish me luck tomorrow!

Brgds, Bert
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 10, 2012 at 05:46 AM.
Reply With Quote
Old Oct 09, 2012, 05:37 PM
Love my scale Whirlybirds
Keyrigger's Avatar
Mississauga, Ont., Can.
Joined Sep 2009
2,380 Posts
Hope the testing does work out well for you. It is rather interesting that the wire broke in the same area as the first one did. As small as the difference is, it may be just enough to alter the dynamics of that shaft. I will certainly keep watch to see how you overcome the issue as I am very interested to see if I am holding back on certain designs (600 size Huey using a flexible shaft rather than torque tube or belt) because of my bad luck with flexible shafts used in model boats. I could never keep the set screws secured on the twisted wire and eventually got rid of my boats. Take care.

Don
Keyrigger is online now Find More Posts by Keyrigger
Reply With Quote
Old Oct 10, 2012, 02:18 AM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Actually, all four of them broke in the same location, i only got a little better endurance by re-aligning. but 3 flights of 8 minutes with one shaft was the best I achieved so far
While fitting the new bearing seat, I noticed that the small inevitable misalignments act like a very miniature hammer on the shaft, pushing it downwards.

I really hope that is the cause of breaking the shafts, because if so, the double ball bearing should carry that load, relieving the flex shaft....
It is interesting to see, that those flex cables for speedometers do have a thrust collar on one end as well, but I noticed that is always at the driving end, and in my case the breakage happens close to the driven end. That confuses me a bit, but hey, you have to start somewhere.... So I fitted the bearing at the driven end, closest to the point of breaking....

The parts are just installed, waiting for the glue to set now, testing will be in a few hours.

The issue with the set screws on twisted wire was my main reason to go with the single strand shaft rather than a multi-wire twisted flex-shaft.
On the 450 at least, the torque is relatively small (5 Ncm approx) and I found that 2 opposite M3 flat nose grubscrews are really plenty. They did not even come loose when fitted without threadlocking agent.
For the interested: on a 600 size that torque would be roughly 30 Ncm in static hoover.

Brgds, Bert
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 10, 2012 at 04:59 AM.
Reply With Quote
Old Oct 10, 2012, 04:58 AM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
First flight, and immediately I noticed some remarkable things....

Noise level seems to be lower. Not measured of course, but the noise seems to be smoother, less whining and whirring. I take that as a good omen.

Noise does not change with aspect: previously, the helicopter sounded different when hoovering side in, compared to hoovering tail in. That has dissapeared.
Seems to me, a good omen too.

Power consumption has increased ever so slightly, but that seems logical, there are two extra ballbearings to drag along (entire tail drive, that is front bevelgear, tail box bevelgear and flex shaft now run in 6 ball bearings instead of four), and I forgot to polish this wire before installing it.

Really hope it is OK now, because I am down to my last shaft, and it is a 80 km drive to my nearest "L"HS and back, just for some wire that cost approx 1.30 Euro per meter...
Just cross my fingers and keep testing.... (boring.... pack after pack in static hoover....)

Keep you guys posted!

Brgds, Bert

EDIT: 2nd battery through.... increased power consumption seems to be a breaking in issue: this time it was even slightly better than before the ball bearing conversion

EDIT2:... 6 minutes into the 3rd battery, it snapped again, and it snapped again, in exactly the same location....
So fitting a ballbearing obviously is not the solution, therefore shaft-hammering or misalignment not the cause....
I really don't get it: the point where it breaks is subject to more or less the lowest torque (even if that is only a very small difference) and it is NOT the point of smallest radius.
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 10, 2012 at 06:32 AM.
Reply With Quote
Old Oct 10, 2012, 07:21 AM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Dissapointed, but not discouraged, I took the bloody thing apart for close-up inspection.

I found that possibly I have made a judgement error that might be the next "most probable cause": for the inner tubing of the system, I used some tubing I had at hand, that was 2 mm inner diameter.
The shaft is 1 mm and that is a considerable play.
When the shaft is going around the bend, it lies to the inside of the bend, except where the tube straightens out. There it crosses over to the outside of the bend.
The breakage seems to happen somewhere in that area.

Logically, the next step, is to provide for a tighter fit between shaft and liner.
The tube of the sullivan bowdencable I used as tailrotor control when the belt was still in use, seems to be a perfect fit inside the original lining, and the shaft seems to be able to turn happily in this lining.

Found one more shaft, so up to the next test....

Brgds, Bert

EDIT: Humpfh! That didn't help either.... again at approximately the same flight time (approx 20 minutes) the shaft gave in again....

BUT: this time the point of breakage shifted upwards by almost an inch, which more or less corresponds with the shift of the contactpoint due to a narrower inner tube....
Maybe the distance between end of the bend to the end of the tube has something to do with it, maybe the exit of the bend has to be more gradual.... Fact is, it does not break at the other exit out of the bend....

It seems I am getting closer to the final answer, but for the moment, I am out of material....

Probably after the weekend, I'll have more time, material and more brainsteam, for now I have had it....
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 10, 2012 at 09:04 AM.
Reply With Quote
Old Oct 10, 2012, 12:38 PM
Registered User
USA, CA, Red Bluff
Joined Dec 2007
952 Posts
Bert: I admire your precistance and patience. Hang in there Ed
prop bender is offline Find More Posts by prop bender
Reply With Quote
Old Oct 10, 2012, 12:54 PM
Complete RC Idiot Savant
The Netherlands
Joined Nov 2009
2,469 Posts
Don't worry, I still have some ideas up my sleeve, and I am 100% sure a single strand flexdrive can work, I have seen that with my own eyes in the past.

It is just a matter of finding the factors that make it fail, and dodging them

So far, to tight radius can be ruled out, vibration and movement, misalignment, obviously all not critical.

Deducting what happenen after the various modifications, now I am thinking that it has to do with the distance between the exitpoint of the bend, and the end of the tube:
The tube is obviously not symetrical on both ends of the tube. The wire strand keeps breaking at one place, approximately at the exit of the bend on the short side. Eliminating consequently misalignment, possible movement and excess play, I can only conclude that it has something to do with the shape of the tube:
On the end where it snaps, the end of the bend is more or less the end of the tube. On the other end of the bend, there is a straight pipe guiding the wire after the bend.

So I need to try and incorporate a sufficient length of straight tube on both sides of the bend.
2 or 3 inch should be possible

But that is at least after tomorrow

Brgds, Bert

PS: obviously, I am as much in the dark as most people, just guessing, deducting and checking off all possible causes for this annoying faillure. If anybody has another view, I'd be hapy to hear it....
Brutus1967 is offline Find More Posts by Brutus1967
Last edited by Brutus1967; Oct 11, 2012 at 05:20 AM. Reason: Additional explanation
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion SMARTMODEL-Tail shaft drive and raise tail system 500/600/800 build thread and Q/A kin Scale Helicopters 32 Nov 04, 2012 10:10 AM
Sold Trex 450 metal tail drive gear - $8 shipped Hemikiller Aircraft - Electric - Helis (FS/W) 0 Sep 25, 2012 09:22 AM
Wanted Looking for Align TREX 500 Tail Boom & Tail Boom Case Belt drive Killer007 Aircraft - Electric - Helis (FS/W) 0 Jul 12, 2012 02:48 AM
Found WTT 450 Sport Belt Tail (for Pro) for 450 Pro Tail SYMAWD Aircraft - Electric - Helis (FS/W) 0 Apr 11, 2012 07:28 PM