Recent content by hushish

Hi, I think the copy of the paper is unclear. I think there is a t3 term in D. Sorry for the confusion... BUT, I could use your help with the final derivation of equation (4)-see attached. I know that the particular solution to the 4th order O.D.E should look something like this: y(x) =...

Hi, Try as I might, I cannot understand how equation (3) with the k4 term was derived. In equation (2), w is a function of t. But in the k4 term after equation (3), there is a t2 term in the denominator. Should it not be a t1 only? What am I missing? Please help. Regards, Steve

Hi Henneh, The following boundary conditions definitely do work in FE. Regards,

Hi Bill, Again, thanks for the effort. All your assumptions are correct. The only thing I can add is that Matlab yields the following more-compact result, but I am clueless as to what Int is (it's not integrate): y =...

Hi Bill, Thanks for taking the time to understand the problem. Now you see why, when we add a distributed load q.x into the mix, it complicates things tremendously. I would love to solve the above problem with a q.x added, as well as for the clamped-clamped beam (attached). Then I would be...

Hi Bill, Sorry about that. The boundary conditions have changed in that now rotations and deflections are zero at x=0. Thus y(0)=0 and y'(0)=0. y(L)=0. R is the reaction force at x=0 and is real and not a function of x. All the assumptions you have expressed in Mathematica are correct. I am...

Hi Bill, For some reason, the approach I took (with your help) above seems to work and my answers are now converging. Perhaps you can add some insight into the next problem: If I adjust the boundary conditions, the problem now requires a particular solution: y''[x]/(E*I)+(P+q x)*y[x] =...

Thanks SteamKing, I was hoping that my case fell inside the "trivial" solution side, but it seems like FE is the only answer.

Thanks Baluncore, I am not familiar with electrical field applications, can you point me in the right direction? If it would help, the force can be distributed over a finite circle of radius R. What would the displacement and load distribution look like then? I assume it needs to be of...

Thanks Bill for your time and help;, I will post something here if I find anything.

Hi Bill, The trivial solution will always tend to C1=C2=0 for buckling, as C1 and C2 are irrelevant. It's always the terms inside the bracket that determine bthe critical buckling load. Only once I have an expression for y[x] can I manipulate the boundary conditions to obtain q. I am...

Hi, Can anybody help me withg the following problem: A rectangular plate with points starting from top left corner and going clockwise:: A B C D. Sides CD and DA are simply supported, and a point force F is applied anywhere on the surface. I am looking for the bending stress distribution...

Hi Bill, You are right, C1 does not equal zero. Nonetheless my answers still do not converge to the correct results. Since the buckling equation is a transcendental equation, it cannot be solved explicitly. Nevertheless, the values of q that satisfy the equation can be determined numerically. I...

Hi Bill, Sorry for the inclusion of the E and I term without explanation. The problem is for buckling of a simply supported beam with an intital force P and a distributed load q. Now the answer converges when there is no P term. However, if there is an initial P term, I am left with the...

Bill, Thanks for the info. I plugged the equation into Matlab and got the following solution for y''+q.x.y=0: y =C1*AiryAi(-(q/E/I)^(1/3)*x)+C2*AiryBi(-(q/E/I)^(1/3)*x) From this, I input (-(q/E/I)^(1/3)*x) into the Airy's functions. y(0)=0, therefore C1=0, and y(L)=0, therefore I solve for...