Design: Empirical or by Calculation

[Simon Wright]

A question raised elsewhere on this forum made me reconsider an old favourite of mine regarding design of model aircraft. Namely does anyone in this hobby really design anything other than by size and shape, the mechanical and aerodynamic stuff that is. The design of the aesthetic or a beautiful shape is an art or skill that few can aspire too and not part of my rambling.

My own experience of designing a large model by calculation from root principles soon foundered when I started following Rule Of Thumb (ROT) techniques. These ROT’s combined with a plethora of estimated figures for material strength, joint and other funny factors very quickly made the precision of calculated figures irrelevant and the accuracy questionable.

Precision figures for Cd, Cl, Cm on aerofoils are similarly based in perfect space and are fudged to realise actual conditions. Even calculating Vmin & Vmax, the prime inputs to any strength calculation is full of little 10% margins and deductions.

My latest radio has the benefit of telemetry and I intend to test out the accuracy of a few theoretical calculations, even here the accuracy of the instrumentation imparts yet more deviations.

Thankfully with small mass, limited size (moment arms) and minimal payload requirement we can easily carry massive overdesign factors between 2 and 10 maybe even more. Other than providing some reassurance by numbers, design by calculation seems to be a pretty fruitless exercise compared to the alternative.

I feel that Empirical design is more the norm using ‘That looks About Right’ (TLAR) techniques with a little bit of copy and paste and a bucketful of ROTs.

If successful model ‘A’ has a wing area XX, tail moment YY and tail area ZZ changing the outline shapes should have very little effect on model ‘B’.

Similarly construction techniques tend to follow monkey see monkey do principles with material sizes, D boxes and structures following methods experienced or seen on similar design.

There is nothing wrong with this technique but it can and does lead to failure when pushing the limits of personal knowledge, materials and technology.

So, after my little ramble are you an estimating mathematician or a very capable mimic.

[kc]

Balsa varies so much in weight and strength that calculations are not very useful. Same with glue and paint. How much you use causes a large variation in model weight.

It's not so much rule of thumb that we use it's practical experiment. If it works we use it.

But if anyone can design a better model by calculation then do it and let us see the result!

[McG 6969]

... bueno... well...

... Aïe nowe nothinge, amme fromme Bruselas...

... but maybe an "estimating mimic"??? ...

Chrisos

BRUselas - BElgica / Controla Trafico Regione No Se Controlo

[Colin Leighfield]

Interesting one Simon. I suppose the first attempt I made at my own design at about twelve years of age, beyond a lot of chuck gliders was a built up A1 size glider design , usual balsa and tissue construction using just balsa cement as glue. It flew well enough. Beyond that I designed a fuselage to take a clapped out Mills .75 and fitted it with the wing and tail from my Mercury Martin glider, which flew very well until my lack of expertise in engine bearer support caused them to come loose. I hadn't got a fretsaw so made the bulkhead out of balsa and it wasn't up to it.

I've had an avid interest in aeroplanes for sixty years and have read extensively, so I'm fairly well informed. Also I acquired a PPL Group A in 1985, so got in some full size experience. Over the years I've drawn up a number of models, most of which didn't get built and a few that did. I generally find that I can design planes that fly, though compared to many of the top guys on this forum, what I've done is very little.

However, I have to say that my methods are empirical and are based very much on experience of what has worked before and other people's successful projects. You develop an understanding of the permutations in aerodynamic detail that give various flying characteristics and how to apply them. There's a lot of discussion about the fine detail of different aerofoil sections, but I have to confess that once having decided on flat bottomed, semi or fully symmetrical, thickness/chord ratio, how far back maximum thickness will be and wash-out considerations, I generally draw them myself and they always seem to be fine. Also the experience with Nigel Hawes flat-plate balsa wings on mid-size models does make me wonder sometimes about how critical some issues really are. It does sound good talking about Reynolds Numbers etc, but sometimes it's best to say "hmmmm- - - - " and just grab a pencil.

Clearly from other threads running now there's heightened interest in structural security with some large model designs, particularly if they are going to be displayed in public. With some of those a level of expertise commensurate with full-size design, including stress loading calculations might be appropriate. However, I would support the view that the over-riding need on "very"(?) large models intended for public display is simple static load testing, not necessarily a sheet of calculations. It would not be difficult to specify where the test loads should be applied, up to a weight commensurate with say 3G? I think that most experienced model designers could build to that requirement using empirical design techniques.

So I am basically a mimic, but even those using a scientific approach are still applying their maths to a background of previous experience. In most cases, I am sure that we design and build empirically with a result that is way in excess of the safety minima that could be determined by calculation and computer simulation. In that sense we are inefficient. If you want to consider someone who really does understand how to use engineering principles to absolutely minimise structure weight, reduce power requirements needed to achieve a targeted performance level and sustain structural integrity, look at Simon Chaddock's successful designs.

[Peter Miller]

I am reminded of a little story.The officer in charge of our model club at St Mawgan in the RAF designed a series of superb control ine stunt models.They were really works of art., the mpdel was named Gaisgeach or Warrior. It featured a fuselage carved and hollowed out from two large blocks of balsa to about 1/8" wall thickness.

The aerodynamics had been studied and calculated to the nth degree. Noel had studied aerodynamics at university.

The wing was a study in the minimum wood for the maximum strength and lightness was paramount. The models were about the size being used for .35 cu.in stunters. but flew on Oliver Tigers.

Needless to say they flew superbly

Then one day he needed a quick model so he knocked up a quick, simple model with a plank wing and box fuselage. He told me that it flew far better than his ultra super models.

Another story which is interesting. There was a modeller who flew in competitions in the USA. He kept winning and all the other modellers kept asking him what airforil section he was using. After all top glider men use Seligs and Epplers and other highly developed airfoils and they wondered what airfoil he had found that was so much better.

His reply sickened them all. He used good old flat bottomed Clark Y

I have recently just designed a small powered glider. Out glider enthusiasts with their fancy foamies were staggered at the superb glide of my Kitehawk.

But then...IT Looks About Right

Don't forget that at our sizes things are not quite the same as large aircraft.

Now you can buy an excellent book by Kermode called the Mechanics of |Flight, It will give you all the mathematical formulae that you could ever wish for and you will be able to baffle brains everywhere.

You can also buy a superb book by Kermode call "Flight Without Formulae" This is the same book without the maths. You will learn an awful lot more useful stuff from that book.

Edited By Peter Miller on 26/09/2016 19:08:45

[Peter Miller]

Just a second thoght.

Clayton Folkerts was a farm boy. He built three basic home builts in the 1920s.

He was then asked to desing a two seat cabin aircraft. He drew this out on the workshop floor.

That was the First Monocoupe which was developed into racing aircraft with such success that they nicknamed the National Air Races the National Monocoupe races. It ws also developed into an amzing aerobatic aircraft and one company was going to reproduce it n the late 90s and di build a prototype.

Folkets also designed a series of successful racing aircraft with the same basic principle. But that was a very common method among racing designers in the 30s.

[Don Fry]

You can't calculate wooden aircraft. An engineer has no data for any of the potential resistance of the material to stress and load and stain. And we havn't started on glue.

An aside, my local fire service used hard wood, ( liga vitae, spelling suspect) wheel chocks. Engineers insisted on plastic chocks, with measurable strength. The service were then sued when one poor person was hit by a high speed 15 kilo lump of plastic spat out of the wheel. My informant reported that everyone knew that wood deformed, and did not spit out. It's a balance of judgement.

See posts on large aircraft current on the site. A bag of worms.

[Simon Wright]

The point I was making on calculation is that, for model aviation it is so full of funny factors and margins as to become almost useless.

Calculations using wood and adhesives are easy, values and factors for the allowable stresses of wood are readily available, the information is published in various locations for most woods and densities, as are the jointing factors for adhesives. Unfortunately the spruce/balsa/ply we use is of inconsistent quality and ungraded.

Calculations for our typical wings with I or C beam spars , boxed with D sheeting tend to ignore the monocoque formed instead focusing on the spars and shear webbing. this builds in another margin for error and over sizing.

The comments on high technology wing sections (aerofoils) raises another topic that leaves me smiling, which is the ability of the average or even highly skilled modeller to achieve a micron perfect section with a knife and sanding block. If the section is close there are very few aerofoils that allow for the sag of covering between wing bays. CNC moulds and composite wings anyone?

Regarding the theory of aerodynamics and mechanical engineering I have a Degree of knowledge in both and while neither a structural or FEA engineer have a good grasp on the  complex interactions, maths and forces involved.  Basically if you build it and it flys without breaking up is it safe, and would some numbers on a piece of paper make it any safer?

 

 

 

Edited By Simon Wright on 26/09/2016 20:44:41

[Geoff S]

When we last visited the USA in 1996 we visited Kitty Hawk and the museum celebrating the Wright brothers achievements. I fully expected it to be full of US bravado and hype but was pleasantly surprised to find it understated and staffed mostly by a band of enthusiastic volunteers. What showed up most was how thoroughly the brothers researched and developed their flying machines. It seemed to me that most of it was using the TLAR principle based on experience. They didn't have a lot to go on but benefited from the collective efforts of other experimenters. I don't think a lot was based on formal maths (or in their case math)

Their machines weren't much more than our big models. Modern designers of full scale aircraft don't have the privilege of using suck it and see engineering not based on careful calculation, fortunately model aircraft designers can keep to the Wright's methods.

Geoff

[Tim Hooper]

Designing a scale model is dead easy, as the designer of the full-sized aeroplane has already done all the hard work for you!

tim

[PatMc]

Posted by Geoff Sleath on 26/09/2016 20:33:16:

When we last visited the USA in 1996 we visited Kitty Hawk and the museum celebrating the Wright brothers achievements. I fully expected it to be full of US bravado and hype but was pleasantly surprised to find it understated and staffed mostly by a band of enthusiastic volunteers. What showed up most was how thoroughly the brothers researched and developed their flying machines. It seemed to me that most of it was using the TLAR principle based on experience. They didn't have a lot to go on but benefited from the collective efforts of other experimenters. I don't think a lot was based on formal maths (or in their case math)

Their machines weren't much more than our big models. Modern designers of full scale aircraft don't have the privilege of using suck it and see engineering not based on careful calculation, fortunately model aircraft designers can keep to the Wright's methods.

Geoff

I think that the Wright brothers were extremely methodical and used formal math[s] in the initial design stages but if necessary reconsidered the data obtained in the light of practical test flying.

Actually the brothers used the published calculations of Lilienthal & other early pioneers for their own early glider and kite flights. However over their first two years of experimental flying [1899 - 1901] they had growing doubts regarding the accuracy of this data so decided to ignore it and rely on their own investigations.

In later years during an interview Orville was asked what had made them successful where the other aviation pioneers had failed.
His reply was simply that "We had a wind tunnel."

 

Edited By PatMc on 26/09/2016 22:56:03

[Colin Leighfield]

Tim, why did you have to blow the whistle? Darn it!

[Peter Miller]

Posted by Tim Hooper on 26/09/2016 21:51:47:

Designing a scale model is dead easy, as the designer of the full-sized aeroplane has already done all the hard work for you!

tim

That doesn't always work!!!!

I never got this one to fly.

[Ernie]

Donald is right, As an architect, I can calculate many structures, because there are readily available figures that tell me about the qualities of available man made materials. Timber is a wee bit more difficult, but for structural use, there are stress graded varieties

Balsa especially is difficult to put a number to. There are so many variables. The same applies to adhesives, and ply to a lesser extent. Also our quality control is very variable

So, off course, we make it all up as we go along, usually making things way oversized and much stronger than necessary, except of course the fins on ultra large scale jets

ernie

[Peter Miller]

It wasn't just the fin, the fuselage disintegrated in mid air as well. Looked as if it had been built from depron and little else.

[Geoff S]

Posted by Peter Miller on 27/09/2016 08:27:44:

Posted by Tim Hooper on 26/09/2016 21:51:47:

Designing a scale model is dead easy, as the designer of the full-sized aeroplane has already done all the hard work for you!

tim

That doesn't always work!!!!

I never got this one to fly.

You know what, Peter? I'm not entirely surprised

Geoff

[Simon Chaddock]

I think it would be fair to say that the Wrights (along with most of early aviation pioneers) had problems understanding aerodynamics (lift to drag ) and control rather than the basic structural issues. The 'strut and wire bracing' construction was quite well under stood.

Then of course there was the problem of the power to weight of the propulsion system.

The Wrights tackled all these issues very methodically and carefully testing as they went along.

Today having sufficient power is not so much of a problem, although I note how much more difficult it is when a particularly inefficient propulsion system is used on purpose - i.e. an EDF!

If you retain the full size construction and material then the structural scale factors tend to work to the models advantage. The problems start when you completely change the material and the way the plane is built.

This was the issue I had when I started designing load bearing structures in Depron. Its characteristics are so completely different that my previous 'wood' experience simply was not relevant.

The only way was to evolve some simple 'rule of thumb' structural tests to avoid catastrophic failures, which so far, has been pretty successful.

[kc]

Andy Lennon wrote varoius books and articles in Model Airplane News which have a more theoretical approach to model design. (I think he was involved in full size aircraft design ) Well worth reading if you can get hold of these out of print items. His designs were mostly plans in M.A.N. and are still available but I don't believe they are generally considered much better than other designers work.despite the theory.

[Jon H]

Interesting thread this as i am currently in the middle of designing my first model from the ground up.

The model is a 90 inch span Petlyakov Pe2 for twin 80 4 strokes and through the whole design i have been actively trying to design it in such a way that loads are transmitted from the engines/landing gear to fuselage and not into the foam wings.

But, while i have been thinking about it and all the rest, the final choices of wood size etc are based upon my experience with other models with zero numerical data to back it up as i dont have any data to work from in the first place.

I do recall some of my mechanics and flight principals from uni, and could work out youngs modulus and all that if i chose, but i dont think it would actually help. My choice of aerofoil was again based upon seeing what the full size used, and then after some hunting found that the section looked pretty good in terms of thickness/camber etc and, it looked right. So that was that (the section is RAF34 with 120% thickness for those who care).

I am concerned about its integrity, probably more so than i should be, but its very much on my mind when choosing things and i might even make a dummy wing for some static load testing, although in truth i am more concerned about the tail with its twin fins than the wings.

[Peter Miller]

Posted by Geoff Sleath on 27/09/2016 11:42:59:

Posted by Peter Miller on 27/09/2016 08:27:44:

Posted by Tim Hooper on 26/09/2016 21:51:47:

Designing a scale model is dead easy, as the designer of the full-sized aeroplane has already done all the hard work for you!

tim

That doesn't always work!!!!

I never got this one to fly.

You know what, Peter? I'm not entirely surprised

Geoff

The take off was the problem, it kept tumbling over No one had the guts to hand launch it!!!

[Martyn K]

I have design my last 3 gliders by calculation (using readily available design tools). Not structurally, just aerodynamically. Thoroughly enjoyed the exercise as well.

Cheshire Cat

Mad Hatter Just started this one

White Rabbit Next winter project

Although very specialist, my models are slightly out of the norm having longer moment arms than is fashionable

Martyn

[Peter Miller]

Posted by Jon Harper - Laser Engines on 27/09/2016 13:52:25:

Interesting thread this as i am currently in the middle of designing my first model from the ground up.

The model is a 90 inch span Petlyakov Pe2 for twin 80 4 strokes and through the whole design i have been actively trying to design it in such a way that loads are transmitted from the engines/landing gear to fuselage and not into the foam wings.

But, while i have been thinking about it and all the rest, the final choices of wood size etc are based upon my experience with other models with zero numerical data to back it up as i dont have any data to work from in the first place.

I do recall some of my mechanics and flight principals from uni, and could work out youngs modulus and all that if i chose, but i dont think it would actually help. My choice of aerofoil was again based upon seeing what the full size used, and then after some hunting found that the section looked pretty good in terms of thickness/camber etc and, it looked right. So that was that (the section is RAF34 with 120% thickness for those who care).

I am concerned about its integrity, probably more so than i should be, but its very much on my mind when choosing things and i might even make a dummy wing for some static load testing, although in truth i am more concerned about the tail with its twin fins than the wings.

I do hope that you mean 12% thickness

Half the secret is transmitting the loads with no sudden changes of section or strength.

Nacelles to spar, spar to centre section.

The skin may carry some loads on stressed skin structures but in models the spar is probaby the main load bearer with a balsa D box LE takinf a lot of it on models.

[Jon H]

Nope, 120%

But thats 120% of the original design thickness as opposed to a 120% overall thickness!

Go here and change the thickness to 120% and you will see what i mean http://airfoiltools.com/plotter/index

Edited By Jon Harper - Laser Engines on 27/09/2016 15:07:40

[Peter Miller]

Posted by Jon Harper - Laser Engines on 27/09/2016 15:06:17:

Nope, 120%

But thats 120% of the original design thickness as opposed to a 120% overall thickness!

Go here and change the thickness to 120% and you will see what i mean **LINK**

Edited By Jon Harper - Laser Engines on 27/09/2016 15:07:40

Ah I see.

I use compufoil for wing designing. Superb program.

[Jon H]

I was recommended that but couldn't make head nor tail of their website. That said, I didn't try very hard as I was already satisfied with what I had. I will make a greater effort next time