I’m developing composite aircraft. We do all our own design, engineering, production.
We use NEiNASTRAN + FEMAP for analysis and VisualMillPro for CAM.
Our plugs are machined on a large cnc, they’re extremely rigid, with a surface out of epoxy-fiberglass. The plug sizes vary from tiny things to a 6.1m long fuselage.
The parts are all vacuum infused using carbon/basalt/epoxy/kevlar and a variety of cores, from Soric-LRC to Diab to COMPAXX-80 from dow.
We are currently at a stage of shooting duratec coats to shine some large parts. The small parts are already done and have a nice finish on them.
Well, that should give some flavor of what we’re upto.
Welcome to the forum. For my finite element class I evaluated a composite model as using Femap and Nastran as well. Do you use some of the advanced modules such as Simulayt and MCT modeling? If so how do like the results?
I see that you use NX, we use NEi. I don’t have MCT, as we have some internally developed script which do similar things, besides we don’t need to pay for a generic MCT which would be useful, but we can still not justify the expense. I’ve decided to spend on our own materials database, this is very difficult to get. Unless the commercial database was generated that replicated all your variables, its just as good as taking data from the still incomplete MilHdbk-17.
We do have the Explicit license, this allows us to do crash simulations such as bird strikes. I believe, in any discrete analysis its the quality of the elements that will eventually give you high accuracy. Take a look at the limit load impact analysis of our composite landing gear spring here: www.nashero.com/Home/Progress
Thanks for sharing your analysis. I would change the unstructured grid on the wingtips to structured and break the surface in global plies to get all the joggles, ribs, spars, shearwebs in place.
Astute eye. You’re right that analysis was NX, but I did use NEi which yielded identical results. This model was for classroom purposes, to demonstrate we knew the basics of the software and solver.
The ply structure was ordered by mapped subdivisions, at the end though the complex geometry made it difficult. For a final class project however it didn’t merit sufficient time to correct, especially since I was just loading the wing as a cantilever beam, meaning my area of interest was on the opposite side of the terrible mesh.
I should mention more clearly;
This simulation was the final project for one my university classes. The class was an introduction on finite element, where I learned the underlying math of FEA and I learned some Femap+NEi. I’m currently finishing up my university studies, hence my forum signature.
NEi was used in the computer labs at school but I used NX at home for a specific reason. PM me if you’d like to know
And lastly If I worked at a company and they approved a design on a model as terrible as this, I would quit immediately!
Good to know that you’re taking a university class on FEA. I’m curious to know what texts do you follow now a days? I did a class in 1985 on FEA and we used a book by O C Zienkiewicz, and the class on Composites used: Analysis and Performance of Fiber Composites by Agarwal and Broutman, the course was taught to us by Agarwal. I still do take some updates now and then, like the workshops that are organized by Prof Steven Tsai (you know the Tsai-Wu criteria). I’d recommend the Tsai workshop to all. Its a 36 hour course and is focused on Composite design-analysis and failure analysis. Especially for current students, since as long as you’re a student, you can register for them for free and they’re held once every 6 months. If you’re a student, and you can get course credits from Stanford Univ for the course.
I’m currently very happy with NEi, besides NX-NASTRAN uses the elements developed by NEi so they’re mostly identical in operation. Siemens also has the $$s to purchase other additions to the NX which they sometimes do, but NX as an integrated system is far too expensive for the features they provide. In today’s times software pricing, the way it is, doesn’t make sense.
Anyhow, I’ve got to work. I’m preparing to shoot duratec on some large parts on Monday and need to concentrate on real things rather than on analysis. Today, I’m also doing a very thick infusion, like 90 plies of basalt, to see if there could be exotherm problems. Wish me luck
