# Rules of thumb in structural design?

### Help Support HomeBuiltAirplanes.com:

#### karoliina.t.salminen

##### Well-Known Member
I have been concentrating on aerodynamics for years and have
avoided the structural design part quite completely. However
now I would like to get started on it too. Before going to complicated
stuff I would like to understand the relations of length, mass and force
(as generated by drag due to and lift).

So a very basic question in roughly rule of thumb category:
- we have sailplane wing that is 25 meters long with AR=25.
Sailplane has vne at 200 km/h and best L/D at 80 km/h and
the mass of the plane is 1000 kg. This is imaginary example, not
a spec of some plane.
-> it is cut into two and it results a wing that is 12.5m long and AR 12.5
If mass stays 1000 what would be the Vne?
How much mass could be carried with the shorter wing
before exceeding the twisting moment or bending limits?

When I was studying, the understanding part (ie rule of thumb) was
lacking and it was putting numbers into equations or integrating
something or stuff like that but a more simple relations probably
exist and I would like to understand those, otherwise I would
blindly trust whatever some excel outputs and believe it to be
right if I could not verify if that is roughly correct and I have learned
to not blindly trust anything I compute without checking if
the answer is anyhow realistic.

#### autoreply

##### Well-Known Member
To answer your question, you need the equivalent of 4-5 years of hard and full-time study.

Rules of thumb are fine, but cutting corners without understanding does not work. Or well, usually once only.

Look up classical beam theory for ratio's and keep the input in variables and see for yourself what aspect ratio, or slenderness does to stiffness, strength etc.

Don't forget about buckling, this (not strength) is the majority of structural work in GA!

Flutter is a topic where you either buckle up for some serious study, or stay far away from it.

Save beam theory, none of the above can be simplified much.

#### karoliina.t.salminen

##### Well-Known Member
So your answer is that the problem is irreducible into a human ubderstandable
relationship, but only analysis of a black box is to be trusted, and hence understanding the phenonena is an impossibility but filling correct equations with correct numbers gives results and this van be completed with 4-5 years of studies. Sure I know that that 4-5 years in university will augment and refresh my forgotten memories of mechanical engineering that I have never needed in my work. However for my skillset it would be waste of time, if I am ever nuts
enough for doing PhD, certainly it is not going to be wing structures. Any other answers?

#### BJC

##### Well-Known Member
HBA Supporter
Refer to the thread on technical references. There are lots of options. My suggestion is to get the text below as a starting point. After some study, you will be able to ask more focused questions, the answers to which will start you down the path of discovery.

Title: Aircraft Structures
Author(s): David J. Peery
Date published / copyright: 1950
Publisher: McGraw-Hill Book Company
Language: English
ISBN: 007049195X

BJC

#### karoliina.t.salminen

##### Well-Known Member
I might have that book in my bookshelf already. Will check.

#### karoliina.t.salminen

##### Well-Known Member
And also is there any good publication for composite structures?
Ie making decisions on wing internal structure (ribs/no ribs/rib spacing/one spar/two spars/layup schedule/ply orientations/sandwhich/no sandwhich/place drill hole ok/place drill hole not ok practical stuff, a designer (=generalist) needs to
understand enough of everything even if specifics were analysed by people
specialized in the given focus area).

#### autoreply

##### Well-Known Member
So your answer is that the problem is irreducible into a human ubderstandable
relationship, but only analysis of a black box is to be trusted, and hence understanding the phenonena is an impossibility but filling correct equations with correct numbers gives results and this van be completed with 4-5 years of studies. Sure I know that that 4-5 years in university will augment and refresh my forgotten memories of mechanical engineering that I have never needed in my work. However for my skillset it would be waste of time, if I am ever nuts
enough for doing PhD, certainly it is not going to be wing structures. Any other answers?
Trusting analysis of a black box should not be done either, ever.

Unless you do that study, you maybe won't comprehend how complex the questions you're asking really are, nor how many different topics you have to study to give a fullfilling answer.

I do this kind of work as a hired gun. The question in your first post would easily warrant a few hundred hours of work, just to do a basic trade-off. That's after studying more than 4-5 years
Anybody who'll tell you it can be done much faster is cutting a lot of corners.

The next best thing is informed opinion, you can get for free on HBA if you screen for the "informed" part yourself ;-)

And also is there any good publication for composite structures?
Ie making decisions on wing internal structure (ribs/no ribs/rib spacing/one spar/two spars/layup schedule/ply orientations/sandwhich/no sandwhich/place drill hole ok/place drill hole not ok practical stuff, a designer (=generalist) needs to
understand enough of everything even if specifics were analysed by people
specialized in the given focus area).
The composites FAQ covers all that. No matter your liking, from simple "how to make" tutorials to deep books that cover stiffness matrices for a composite layup and how that influences eigenfrequency.

http://www.homebuiltairplanes.com/forums/composites/8888-read-first-composites-faq.html

#### karoliina.t.salminen

##### Well-Known Member
yeah if I can get into that with enough financial basis one day, I may hire someone
like you to do analysis. But the analysis is not for cutting wing half as
spending hundreds of hours to find answer to that question does not
have value because it was an imaginary example. And in fact in my
current scale it is less expensive to build a test piece and test it to destruction
than e. 800 hours of analysis for what happens if is 20 cm longer.

#### Topaz

##### Super Moderator
Staff member
Log Member
So your answer is that the problem is irreducible into a human ubderstandable
relationship, but only analysis of a black box is to be trusted, and hence understanding the phenonena is an impossibility but filling correct equations with correct numbers gives results and this van be completed with 4-5 years of studies. ...
I don't think he's quite saying that at all. Prior to the mid-1960's, all parts of every airplane structure (or any structure, for that matter) were calculated by hand, with at best a slide rule and mechanical calculator (what we, today, would call an adding machine). My understanding is that even the Mach 3+ Lockheed A-12, the airplane that later became the SR-71 Blackbird, was designed completely in this this way. So no, of course, no "black box" is required to do structural analysis. Honestly, it's my personal belief that it you can't do it by hand, you shouldn't be using a computer for the work, either. You have no way to judge or check the computer's results.

That said, structures is an immense body of knowledge. I'm pretty much exactly where you are, Karoliina: Strong on aero and stability/control, and weak in structures. Aside from some decent procedural knowledge in building with Rutan-style moldless composites, I have a lot to learn about designing a structure that I can feel is safe and properly light for the loads being applied. I'll be on this same road with you very shortly. I've done some early study, and have at least the shaggy outlines of the learning task staked out.

The best resource I can suggest is contained in the reference listing compiled by our late member Orion, and specifically in post #3. There is a lot of material there and you can focus on particular technologies if you like, since he's categorized the topics by building material.

Will it take you 4-5 years of study? As is often the case, I think that depends on the definition of the task. If you want to become a competent structural engineer to the level of being able to take the topic professionally, across multiple materials and aircraft types from ultralights to supersonic jets, then yes, 4-5 years is probably accurate. On the other hand, most homebuilders specialize, and I personally don't believe that it would take nearly as long to become "good enough" to learn the fundamentals, and how to safely analyze and design structures in a given single material choice, for the kind of airplanes we talk about here. While a lot of the knowledge "crosses materials", there is a lot of material-specific knowledge, too.

So, TL;DR: No, there really aren't any "rules of thumb" for structural design. However, learning that field of study to the level needed by a homebuilder need not be a life sentence. Look to Orion's listing for some really good resources to help you on your way.

#### karoliina.t.salminen

##### Well-Known Member
Thanks for guidances and sympathies. Indeed I get it, it is a multidimensional
optimization problem and there is no one-dimensional answer aka rule of thumb. Will check Orion's list. Sad that he is not here anymore.

Tom, ref. tyranny of the experts, (weak) AI will replace those experts first that are
sitting on an equation or a set of equations and data. As I noted, high dimensional optimization problem that is not well suitable for human computation as tedious but optimal space exists and should be fully computable.

#### autoreply

##### Well-Known Member
Why not put up a concrete concept, some specs and a goal? Because given the direction you were (and are) heading, I doubt it's much of an issue in the first place.

#### karoliina.t.salminen

##### Well-Known Member
I will start a new thread when I have specs and mission profile set.
Currently they are not for full scale. For UAV I know what one customer would
need and specs for that are sort of known and it would be applicable for
other similar cases as well. To test stuff for full scale, the sub scale will have the same configuration.

The original spec for full scale still exists: >=5000 km range, 4 people with survival equipment, reasonable time to travel the 5000 km, slow stall speed, slow landing speed (slower stall than DA40). But it is a new topic sometime later. I need to check some distances on map, maybe even 5000 km is not enough - circling the Earth to east from here would need passing some questionable territories without landing for it to be doable and this sets the range parameter.

#### Pops

##### Well-Known Member
HBA Supporter
Log Member
First rule of thumb in structural design is--- There is no rule of thumb in structural design.

Been an airport bum since 5 years old and was designing my model airplanes at 10 years old, and after 75 years, I'm beginning to learn how little I know. ( I'll say what you are thinking, "Must be a slow learner". Maybe, but its been fun.

Dan

#### WK95

##### Well-Known Member
There are rules of thumb but they should only be used as first-order approximations during conceptual design.

One such rule of thumb is the 10% rule for composites where

Simple, reliable methods are presented for calculating the fiber-dominated in-plane strengths of well designed fibrous composite laminates. This is accomplished by a simple rule of mixtures, the Ten Percent Rule. The primary fibers for each uniaxial load condition are considered to develop 100 percent of the reference strength of the composite material for each environment, while the secondary (transverse) fibers are credited with only 10 percent of this strength and stiffness, whether they be inclined at 90' to the primary fibers or at ?45 deg.
Source: https://www.sawe.org/papers/2054 and Raymer's book

Don't expect accuracy but usually, you'll get ball park results. In fact, the entirety of conceptual design relies on quick and dirty calculations some of which are really just based on rule of thumbs.

#### WonderousMountain

##### Well-Known Member
Simplified equations for generalized results.

Can't say the sun is shining.

We have to say the light is reaching it's destination.

Now you know everything you need to know about engineers.

#### Jimstix

##### Well-Known Member
Ah, rules of thumb (ROTs) for the structural designer. Here are some ROTs that I find useful;

1) When in doubt, make it stout.
2) If it is made for flight, make it light.
3) Loads are carried first by the stiffest part.
4) Avoid small diameter bolts.
5) The shear center of a structure is never coincident with the load center.
6) Assume that the wing spar(s) carry all of the bending loads - until proven otherwise.
7) Stress analysis is easy, loads analysis is hard

I hope that these hard-learned ROTs can help you in your quest. Jim

#### DangerZone

##### Well-Known Member
HBA Supporter
I might have that book in my bookshelf already. Will check.
Bear in mind the 1950s version which BJC recommended is better than the newer one written somewhere around the late 1980s (or early 90s?) which was supposed to be a second edition but turned to be worse than the original. The original is really 'old school', easily understandable and quite worthy in logic thinking about designing aircraft structures.

And also is there any good publication for composite structures?
Ie making decisions on wing internal structure (ribs/no ribs/rib spacing/one spar/two spars/layup schedule/ply orientations/sandwhich/no sandwhich/place drill hole ok/place drill hole not ok practical stuff, a designer (=generalist) needs to
understand enough of everything
even if specifics were analysed by people
specialized in the given focus area).
Not many, there was one coming from a European initiative to help aerospace students in this but it was quite elementary. Your airplane is a glider (judging by a large wingspan) and tips&tricks for this kind of aircraft might differ from a fast racer design. The thing is that most of your bold questions are the ones that should be left to the designer's imagination and intuition. You will have to decide yourself which technique to use according to what you will see best for your mission requrements. The best rule of thumb is: get the most/maximum out of least effort/time. Deisng with one wingspar instead of two because it saves time and it might possibly be better than the two spar weight wise. Don't drill holes if you can use a better technique of bonding. Check advice in books like those that Bingelis wrote, pure practical stuff with pros and cons. And test everything before use because this is better than only relying on books or publications. Gliders are not my field but probably most of these ideas could apply to them too, to keep everything as simple as possible.

#### plncraze

##### Well-Known Member
HBA Supporter
Get a copy of WJ File's book from the EAA book. It is all "rules of thumb" or as the title says "practical" design advice. There are other practical design books out there. Many aircraft design books from the thirties are like this. Also check out Richard Hiscock's design book.