Hand Stress Analysis 7

[prost]

Not to disparage anybody's quality of training in aircraft structural analysis, nevertheless I have noticed a recent trend I thought this group could help explain. Just perusing the many ads for "stress analysts" or some similar description, I have noticed that lately in addition to asking for experience with "FEA" that "hand stress analysis" experience is also requested. It might be that my training is substantially different from that of others, so I don't know--don't all FEA stress analysts know how to do hand analysis? It would seem almost obvious to me that you would need to be able to do hand stress analysis before moving on to FEA, and that hand stress analysis is still a routine part of any finite element analysis of a structure. What am I missing?

 

[prost]

Just a prediction, crackman: corporations will ship every engineering job out of country before they commit to something even closed to apprenticeships. I like the idea of somehow making the ivory towers accountable; that might actually be worth pursuing, though I suppose if the corporations don't demand it, nothing will happen.

What would cause this issue to 'come to a head'? Catastrophic events such as the Hyatt pedestrian bridge or any one of a hundred plane crashes don't seem to cause these issues to come to head, what would?

Back to the original subject, 'hand stress analysis'--don't they teach that anymore at universities?

 

[IRstuff]

Good thing I'm in defense. Our latest unclassified contract requirements read like a special access classified program. Can't even talk to foreign persons, in or out of the US, much less farm the work out.

TTFN

 

[wes616]

IR,

I always thought Aero would be like that, but after experienceing the outsourcing on JSF I don't think that it'll be too long before ITAR is pretty much meaningless.



Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.

 

[IRstuff]

JSF is indeed a different animal, since it was pushed through the funding cycle with the expectation that Europeans would partake in the feeding trough, specifically so that the Europeans would put their joint fighter on the back burner.

TTFN

 

[Sparweb]

This thread could be re-named "Training New Hires, Part 2".

I do agree, Prost, that without a solid hand-calc background an engineer cannot understand the stress model he sees on the computer screen. But that experience is only going to come from us, and only if we take the time to do it.

Getting back to Crackman's comment:

"Universities of today only provide the basis for building an engineering career but not the tools to actually perform the task."

I don't think there's much we can do about the graduates produced by universities. Experienced senior engineers must cultivate a mentor-apprentice culture in their respective organizations, and resist pressure to the contrary. We may be bad as teachers, and often hesitate to stop for a moment to explain the meaning of what we're doing, but I don't see much choice. We would all be the better for it, too.

Stories of junior engineers laid off by management, graduates who can't draw the correct FBD, fatally flawed FEM's, etc. may be discouraging, but that is no excuse to throw up our hands. Quite the opposite; it is telling us what we have to do.

Electricians, airframe & powerplant mechanics, and carpenters, they all must spend time as apprentices to learn the craft and earn the ticket. Are the trades so far beneath us lofty engineers, that we cannot use a similar system (even if informally?). Doctors do residences before surgeries, and academics do post-docs before teaching. It's not like we don't have any practical examples to learn from.

I would add that I graduated from a fantastic college program that strongly emphasized analytical methods of aircraft structures. I came out of school having pounded dozens of rivets, read a dozen chapters of Bruhn, and cranked out dozens of detailed drawings. Seems it's a rarity. Since the program was cancelled by the school a few years ago, the aircraft industry in Canada is the poorer.

I know I was lucky, but most don't get such a good start. It's encumbent upon us to bridge the gap.

One last point, since I'm on a roll:

Consider how long it takes to train anyone to work well and independently as a aeronautical engineer, even if he has a good mentor. Unless he gets insanely specialized, it can take over 5 years (after appropriate schooling) for him to be ready to do his own design projects without constant supervision. Think of all the drafting, sizing, communicating, testing, negotiating, and analyzing that has to be done, and none of it can be learned overnight. Only a fraction can be learned at school.


Steven Fahey, CET

 

[plasgears]

I have used computer analysis extensively, but I sometimes back it up with hand analysis to generally check for order of magnitude. Consider that some programs are written by computer nerds with little or no design and field experience. Some spectacular failures have been blamed on this lack of detail "feel" in computer designing.

Another factor is that semi-professionals are pushed into computer analysis. In Michigan wooden roof trusses are common in construction, and we have had some spectacular roof failures, even in state buildings.

 

[thruthefence]

First of all, I'd like to state that maybe I'm out of my element here, as I am "just" an Aircraft Mechanic, although one with 35 years of experience on Rotorcraft & part 25 aircraft.But, over the years I see the same mistakes made over & over again in newly introduced aircraft. I will give one example: Transducers & Pressure switches , such as oil pressure, torque, fuel pressure, etc come from the mfg. mounted directly to the powerplant or transmission case. They run a few hundred hours & fail, and after a time, the mfg issues a SB to remote re-mount the component using adel clamps & a flex hose. I've seen it on many different aircraft. Sometimes, as an interim, they'll introduce a more robust switch, ( at triple the cost ) but sooner or later,we usually end up clamped up at the firewall. Is this the result of simplicity of design at all costs, without looking how this stuff works in the real world? Thanks for listening. Steve

 

[IRstuff]

No, it's probably a lack of data followed by a component manufacturer who doesn't do extensive vibration testing.

The design requirement flowdown for the transducer would include the vibration profile of the mounting. And generally, the component in question can indeed operate in such an environment. However, fatigue failure data is usually very sparse, so an engineer would ask for data from the manufacturer and not get any and march forward because there's no red flag.

TTFN

 

[HANDCALC]

Prost,

Since my moniker indicates a bias for hand-calculations, I have to add another perspective as to why ads for "stress analysts" would explicitly request experience in hand-stress analysis.

My generation started out with the slide-rule in High School, progressed to calculators and Hollerith punch-cards in College and Universty and (if lucky) we gained limited experience with FEA packages in industry. Stress reports and aircraft drawings were prepared with a pen or pencil, and right or wrong, engineers were expected to perform simple stress checks to ensure parts were strong enough for their intended purpose. Detailed stressing was passed to more experienced engineers (if required).

Today's generation are better equipped to use computers to design, analyze and draw components, and they accomplish these tasks much quicker than the previous generation. What I have observed, however, is that younger engineers and designers often lack the ability to check their own work. Frequently I am asked, "Is this strong enough?", and my reply is "How strong does it have to be?" Then I use the opportunity to teach about loads, material properties, manufacturing techniques, Free-Body Diagrams and design (airworthiness) requirements.

Depending on the age of the person who wrote the job ad, they may need someone who is equipped with other tools to check their work. Not all stress analysts have been properly instructed in their trade, and these tasks are sometimes pushed down to design draftsperson who have never been formally trained in stress analysis, and are simply running FEA packages attached to their CAD programs.

Depending on financial constraints, not all organizations can afford the licence fees associated with multiple copies of FEA packages, and many are still using legacy data (hand-calculations) on a daily basis.

All engineers have a responsibility to share what they have learned with the next generation, and vice-versa. I have mentored a new hire in my department with challenging work assignments, have shown him when hand-calculations are sufficiently conservative, and he in-turn has shown great patience in educating me on how quickly he can use computers (and free downloads) to solve more difficult problems. We teach each other...and learn together...

Hand calcs are still needed, and authors like Bruhn, Peery, Niu, Megson, Flabel have left a wealth of information behind for those willing to learn....

 

[crackman]

My basic comments regarding universities is really about them taking ownership with at least exposing students to the tools which the industry uses. Most graduates to this day still come fresh out of school without even having ever heard or seen MMPDS. This I do attribute to a failing of the universities as they are still overly focused on the theory and not providing exposure to the practical side. Just a comparison of the times, look at the Purdue website. They have a picture of Bruhn teaching a class. In the middle of the class is an actual fuselage used as a training tool.

Anyways, back to the hand stress analysis requirements for jobs. I think one of the reasons for this too other than those mentioned by HANDCALC is that when reports get approved it gets very difficult to document computer runs and results. I know for a fact that the initial Raptor stress reports had to be completedly redone because when the USAF reviewed them they were full of nothing but Excel spreadsheet results and found unacceptable. The USAF requirement included hand stress calculations for verification purposes of all computer models, calculations, etc. This is true with the FAA world as well. So, many agencies use the hand stress calculation as a validation tool and therefore companies need to require it.

 

[corus]

In my line (steel) all my work has to be validated in some way or another, not because the structure may fall down (as it usually has done already) but because people won't believe the results. At least the USAF shows some commone sense in seeking validation but I'm a little surprised that the aircraft industry doesn't in general, or so it seems. The next time I fly I shall demand a parachute, or at least wear extremely baggy trousers.

corus

 

[SAITAETGrad]

Interesting debate.

Those of you have not read Petrowski's book "To Engineer is Human" will enjoy the chapter "From Slide Ruler to Computer (or Forgeting How it Used to Be Done).

 

[YoungTurk]

I'm in payloads/interiors, and around here its almost 100% hand stress analysis. The DERs I support would have it no other way. This is true across my entire organization. My engineering mechanics degree (U Wisconsin), which focused on aerospace, prepared me quite well for this.

The above is like reading about some alternate universe to me.

BUT, I used to work in research and my senior ME kept insisting we FEA everything. Never mind early attempts at said FEA produced failure loads 10x the actual measured loads- it was FEA all the way. Seemed he was responding to industry "trends". Have you seen a copy of Machine Design lately? There are consistently multiple articles and many more "product info" sections on FEA in there and also in most similar trade magazines. There's usually an ad for the latest and greatest FEA package on the facing page.

Just a thought here: there is money, BIG money, to be made on FEA by the vendors. Might they be influencing the industry with this type of "information"?

 

[prost]

Thank you all for your inputs. It appears to be the near consensus of the people who chimed in that mechanical and aerospace structural education at the universities in the US is so focused on using FEA to solve all structural analysis problems that hand stress analysis (what most primarily experienced engineers would use for their "nonsense" filter, that is, the filter being some way to check the nonsense that sometimes can be calculated by a computer) is being severly neglected, and that even corporations are recognizing this neglect. We can debate the ethics of the decisions that universities are making when they neglect such tried and true methods as traditional hand stress analysis, nevertheless, it apparent that many newly minted engineers (say with 5 or less years experience) are not receiving necessary instruction in simple hand stress analysis techniques, techniques they will need the rest of their careers, I suspect.

This might be related, it might not be. A number of years back, the accrediting agencies and prof. societies began pushing universities to include a full year of "design" class work for undergraduate engineering students. Since the push to make engineering a 5 year degree has petered out, that means that requiring a full year of anything that isn't there now is a zero sum game--if you suddenly demand 8 more classes in design (vs. the 2 that was in the normal curriculum at the time), and you still need the degree completed in 4 years, what do you delete? I asked some professors about that 5 years ago, and they said they looked at the individual classes and decided there was considerable overlap. I didn't have a clever response at the time, but now thinking it over, I would have asked them this--isn't there quite a bit to be gained by the repetition? For myself, I know that I had to take quite a few math classes, some material from which was repeated in various engineering classes. I also know that such repetition helped me remember that math both for my future graduate degrees and my work. I also know that repetition in hand stress analysis techniques also continues to help me in my current work.

I might just be 'out there--Does anybody else think that the full year of design classes requirement has contributed to the big holes that are apparent in current undergraduate structural engineering curricula? I can imagine other applied mechanics areas are suffering, such as fluid mechanics.

 

[Kwan]

I think for the typical aerospace degree it's a lack in teaching the traditional methods for structure. I graduated in 95 and my structure's classes were mostly FEM. I do remember a little buckling and such, but they didn't go into real world problems like non-linear buckling, diagonal tension, etc. I would have been better off learning the traditional analysis like unit beam method, ring analysis, joint analysis, etc. and not have FEM at all.

A good structure curriculum should have the traditional methods. It's okay to also include FEM. But the FEM should be compared to hand analysis and testing. This way the graduate would have valuable information regarding the strength and weakness of each method.

 

[YoungTurk]

"It appears to be the near consensus of the people who chimed in that mechanical and aerospace structural education at the universities in the US is so focused on using FEA to solve all structural analysis problems that hand stress analysis ...is being severly neglected"

WOW! It appears Kwan agress with you, but are there any other recent graduates who can attest to this, or is this just the group's perception? Was my college experience an abnormality (Eng Mechanics with aero emphasis, about 6-8 stress courses, only one FEA)?

 

[prost]

It wasn't my perception that recent engineering grads were not receiving hand stress analysis instruction until I noticed the unusually high percentage of job ads asking for 'hand stress analysis' experience (something I don't recall even 5 years ago coming up in job ads). My survey wasn't scientific of course; nevertheless my point was that if even employers are asking for 'hand stress analysis' experience (something in the past that wouldn't have been asked because employers assumed that everyone 'knew' how to do hand stress analysis), then there must be a large hand stress analysis void in recent grads' education and experience.

YoungTurk--what university?

 

[pyro]

Graduated in 2000 from University of Missouri-Rolla, we had one FEA course that was almost an afterthought. We didn't even get on the software until halfway through the semester. The course was mainly about applications of matrix algebra to engineering problems with a short section dealing with element types and loads/constraints. The rest of the courses were strictly hand calcs or MATLAB for the more calculation-intensive problems.

 

[JakeLiv]

I've joined a nice aviation company as their sole structural analyst, and had very little hand-calc experience prior to that.
The hand calculations are not that difficult. If you know the basic rules, the stress/stran relations, and how to use the propere literature (bruhn, niu, etc.) you can do a lot of work manually, even without needing to go to the computer.
I have all the software I need, but sometimes it's just faster to give a general result/estimate using hand calculations (to be backed with further analysis).

 

[YoungTurk]

Prost- U of Wisconsin, 2003.

Is it possible that these job postings are making a differentiation between actual on the job experience and class instruction? I could see that being a much smaller group as many of the people I went to school with went on to non-stress related fields, even though the engineering mechanics and astronautics degree I got was oriented toward just that type of job.

Pyro- Glad to see I'm not the only one with similar experience.