Thread Tools
Jan 19, 2012, 06:49 PM
Registered User
What is the key to finding out where the turbulator strip is placed?

Does the layer stay a constant function of the chord or a percentage of the chord? Might the thickness of the turbulator need to be varied?

Sign up now
to remove ads between posts
Jan 19, 2012, 08:51 PM
Registered User
This is a very interesting approach. Way back when I was looking at airfoil/wing optimization I wanted to look at the effects of forced transition to see where that would lead. I'm glad you are applying your time towards this. I think it could offer a lot of promise... at least it needs to be explored.

I think a collection of these wings built by different folks would go far towards the understanding of these prediction/design methods.

Jan 19, 2012, 09:45 PM
Registered User

The location of the desired transition is specified in the name of each foil. It is the number after the "T" in the name. To have the transition occur there, the turbulator needs to be slightly in front of this location. I have not investigated the required thickness or thickness distribution yet. I figured on a zig-zag turbulator for now as that is one of the simplest effective ones. I'm sure if this works out that there will be many people looking into the best turbulation methods. We may even end up looking at active turbulation eventually.

Jan 19, 2012, 09:47 PM
Registered User

Yep, that's why I published this even though the work is incomplete by my standards. I think real world testing is warranted before we go too far down this path.

Jan 19, 2012, 09:53 PM
Registered User

Thank you for the link!

Jan 19, 2012, 10:09 PM
Registered User
Here is what the foils do when no turbulation is present.

Jan 19, 2012, 10:26 PM
Registered User
Here is showing Zone-V2 moldie reference at other camber settings. The possibilities are eye-opening. However, we don't know the drag penalty for the turbulators. So, performance will be less than this at the higher speed end where they are doing their work. How much less we do not yet know.

Jan 19, 2012, 11:52 PM
Registered User
Here is the forgiving series at +6 degrees camber. Call it a float/thermal setting... It looks fairly forgiving.

The ripple in the shape of this curve indicates the turbulator is not quite far enough forward to be optimal for this camber setting. The cruise camber setting (+2 degrees) also shows a slight ripple. It may be beneficial to move the turbulator forward a couple of percent compared to where I specified. Again, this is the sort of thing real world testing should show.

Clearly the response curve for any particular camber setting is quite wide. There is great latitude on what one might call a "cruise" setting or a "float" setting.


PS - To put this data in perspective, fly 1M/s too fast for minimum sink at +6 degrees camber with Zone-V2 or with the forgiving series. The forgiving series gives up 5cm/s (2"/s) less altitude... But again, this assumes zero drag for the turbulators which won't be the case. The shape of the curve indicates that in all likelyhood, one can carry greater camber for the same turbulence level and therefore reduce sink rate that way.
Last edited by G_T; Jan 20, 2012 at 12:52 AM.
Jan 20, 2012, 03:17 AM
Team WC2013 F3K
oakman7004's Avatar
Originally Posted by G_T
Here is showing Zone-V2 moldie reference at other camber settings. The possibilities are eye-opening. However, we don't know the drag penalty for the turbulators. So, performance will be less than this at the higher speed end where they are doing their work. How much less we do not yet know.

Over the years we(RCG) have discussed this topic and I always follow with
big interest since I have used trips towards the tips for many years now.

And just the other week I serached for a statement by Dr Drela were his says that the drag from the tips will in reality not be noticable. Not even in launch...(but I failed to find this statement in my seach efforts)

However, for me this is very much inline with my findings over the years. For me the trips (used on several different airfoils) gives noticable improvements (low speed) and no noticable drawbacks, not even in launch. I am using trips, today, that in theory are to high by 0.1mm the last 10 cm. Last, the improvemets are more noticable with higher wingloadings (like 15-18g/sqrm).

So my conclusion is that the drag increase is so small and other factors are way more important, e.g. fuse area, airfoil. and therefore one cannot notice the different in the increased drag.

/Jonas Ekman
Last edited by oakman7004; Jan 20, 2012 at 03:18 AM. Reason: typo
Jan 20, 2012, 03:42 AM
Registered User
The link below was placed by Benjaminr in another thread. It gives an indication of the drag caused by the turbulator itself.

Jan 20, 2012, 08:34 AM
Aurora Builder
Very interesting paper, written by a prof of mine. Really should go spend some time talking about these issues with him when I can.

Key quote from the paper: "The tripped airfoil does show
an improvement at the high-Cl , lower Reynolds number condition,
but this improvement is compromisedby a small, but important loss
in performance at the low-Cl , higher Reynolds number condition."

Keep in mind low Reynolds number is 100,000 in this study, which is a pretty high Re value in the DLG world. My take, based on the data in Dr. G's study, the data presented here by Gerald, and the evidence from Jonas, is this is a direction worth taking a very serious look at. Will definitely build a 4-5 panel bagged wing, would be great if I had it done for San Felasco but I doubt it will be ready.
Jan 20, 2012, 08:38 AM
Aurora Builder
Oh, something else critical from the article: "Owing to the lack of good airfoil
analysis methods that can predict the effects of trips, airfoils
designed to use trips currently need extensive and systematic windtunnel
experimental investigation as a part of the design process.
For this reason the study highlights the needs for the development
of empirical and computationalmodels that can account for the different
effects of boundary-layer trips during the design stages of a
low Reynolds number airfoil."

Gerald, any thoughts on this? Build and fly seems to be the best approach to this type of work...
Jan 20, 2012, 12:48 PM
Registered User
That's why I published this work in progress. It is far enough along that it needs real world testing.

Admittedly real world testing is difficult at best. Conditions change. Pilot's perceptions change. Pilot's skills differ. How well a pilot sets up a plane varies all over the place. Accuracy of wings varies tremendously. There is a learning curve with any new plane to fly it most efficiently.

However, that is real life. What the preliminary analysis of this series shows is that the difference could potentially be large enough that Stevie Wonder could see it.

Consider this design an example of the sort of wing where synchronization is key and turbulation is used as an important tool rather than as an afterthought.

Jan 20, 2012, 01:27 PM
Registered User
Philip Kolb's Avatar

Some basics on turbulators

Hello Gerald, hello RCGroups community,

I was just following the latest discussion here and reminded about some work I did seven years ago in preparation for the 2005 F3J European Championship.
My plane of choice that days was a samba Model Pike superior which featured a Helmut Quabeck section, the HQW 2.5/8.5.
Knowing that these sections have some issues with excessive bubble drag especially in the low Reynolds number regime I started to calculate and experiment with turbulator strips. My first calculations showed very similar results like the ones shown in Geralds graphs - just for different Reynolds numbers. That made me very happy in the beginning because the amount of drag saved seemed to be significant!
It made me think further and finally re-think.
How could I be so sure about the few results of the Xfoil calculations I made by just fixing the upper surfaces transition point?
How could I trust them without knowing anything about the boundary layer conditions nor anything about how Xfoil is calculating the location of the free transition in accordance to the ncrit numbers chosen?
How could I trust these polars without taking a closer look onto the topic of how laminar seperation bubbles build up in accordance to the form parameters of the airfoils?
What I want to point out with my little story here is, that this is a very complex topic which needs some basic understanding before moving on to calculation or even field testing.
The paper of Dr. Gopalarathnam Bas was posting is brilliant work and leads to a lot of answers concerning the utilization of trips or designing 'tripped' airfoils.
I just want to point out some basics about turbulator strips which need to be considered before starting to design or calculate with them.

- The turbulator strip shall trip the boundary layer to facilitate transition from laminar to turbulent. This is especially favorable in cases where the transition goes ahead with large laminar separation bubbles and thereby excessive dragrise due to them. Tripping the boundary layer before its point of natural transition might result in less drag thereby.

- The point of free transition on the upper surface is moving upstream with increasing alpha (increasing cl). If the turbulator strip is placed upstream of this point it can trip the boundary layer and force transition. If the turbulator strip is placed too far downstream it might either be inside the laminar separation bubble or in fully turbulent reattached flow. If so, it is ineffective in ways of facilitating transition.

- The turbulator strip itself will cause additional pressure drag. It is higher when the strip is positioned in laminar flow and relatively low when positioned in turbulent flow. Nevertheless it will contribute drag to the overall drag of the aircraft.

Sadly none of these points are considered in the graphs shown by Gerald. They might lead towards some misconception for the calculations are only done with forced transition. A turbulator strip thereby must be placed that far upstream of the transition point to still trip the boundary layer upstream of the laminar separation bubble. By solely looking on the diagram of chordwise transition over alpha one can't tell about where the buildup of the laminar separation bubble is about to start. It is even difficult to get a hint about this by taking a look upon the cp distribution for a given cl. The only possible way to get this hint about the startpoint of transition is by taking a look upon the skinfriction coefficients. There is a possibility to plot these skinfriction coefficients in Xfoil. As these coefficients turn negative they signalize reverse flow on the airfoils surface and thereby show the possibilty of separation building up. This point might be further upstream than the transition curve will show it what could later lead into this misconception of saving significant amounts of drag in areas where the turbulator strip will already be placed inside the separation bubble. The effectivity of a turbulator strip placed on (f.e.) the upper surface of an airfoil is only given as long as it is placed upstream of that point!
I attached two pictures to explain a suitable way of finding out about a startlocation for the turbulator strip.
One of the pictures shows polars of the HQW airfoil used on the Pike superior. One polar showing 'Cl over Cd' using free transition, the other using forced transition on the upper side at 0.68 x chord.
The turbulator was meant to be placed that way to safe drag in areas where bubble drag is most excessive - this is around cl = 0.35 (for the given Re*sqrtCl) for that airfoil.
By plotting 'cf over chord' I can find out that cf-values turn negative at about 68% of the chord, the upper side's transitionpoint thereby is further downstream at nearly 92% of the chord.
With this in mind I can start setting the transition at the mentioned 68%.
You can follow this method in the second picture I attached.
After that, one can see the amount of drag saved by using forced transition - the difference between the two Cl/Cd polars (one red, one orange) in the first picture up to Cl = 0.35. Above Cl = 0.35 we know, that the turbulator strip placed at 68% chord ( or even a little upstream of that) would be ineffective because it would already be located inside the laminar separation bubble.
This triplocation will not lead to better performance at Cl-values higher than 0.35 although the polars might suggest this! Additionally we need to consider the 'unknown' amount of pressure drag added by the turbulator strip itself especially at Cl-values below 0.35.
And there you need to be careful reading Geralds polars as well...One should know the design Cl for the triplocation before judging the results first.
As you might see, it is a lot of 'juggling' required to design for proper triplocations or getting better performance out of our planes by utilizing turbulator strips. It nevertheless might be helpful in consideration of the planes handling (f.e. aileron effectivity) or lead to better performance in a desired (mostly narrow) cl-range.
Have a great weekend:

Philip Kolb
Jan 20, 2012, 02:42 PM
Registered User
Hi Philip,

thanks for your post - you displayed a lot of patience! One point I am not happy with is that people trust in calculations without looking at the tools they are using.

I really would like to hear Mark Drela's comment on the designs presented here.


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Airfoil and plane design work for 2012 G_T Hand Launch 171 Dec 28, 2011 07:18 PM
Discussion Will this slightly different mixer set up for an Octo work? jessestack Multirotor Drone Talk 1 Dec 07, 2011 10:45 AM
Discussion DLG Contest Schedules for early 2012 FLY F3B Hand Launch 8 Oct 11, 2011 12:04 PM
Idea Small 30" DLG Design Concept for Spektrum AR6400L Collaborative Work.. RobertTHaas Hand Launch 13 Apr 17, 2011 09:20 PM
Alert Listed my Rex 450 SA for sale... going in a different direction Sky_Dancer Mini Helis 3 Nov 28, 2008 10:06 PM