Comments on: Matlab

By: Pierrot

Pierrot — Wed, 30 Nov 2011 16:27:07 +0000

Sebastian,
a partial triangular ascending load is included in the c++ application, just use a negative value for the load length.
Kind Regards,
Pierrot

By: Admin

Admin — Tue, 20 Sep 2011 17:32:43 +0000

Sebastian,
Your English is very good! I did not include the ability to apply this form of load in the program. Pierrot has included the ascending type only. You can try to include the load in either the Matlab or C++ version yourself if you know programming and the theory (and if you can follow the code). Otherwise you could try a freeware program like LinPro (http://www.line.co.ba) which is really good and easy to use.
Hope that helps,
Colin

By: Sebastian

Sebastian — Tue, 20 Sep 2011 16:26:58 +0000

Hi,
Thanks for your useful site and to share all with us.
I have a quiestion. My english is not the best, I hope that you understand me!
How can I put a partial triangular ascending load? I’m trying in your Matlab program and in the c++ aplication made by Pierrot. In the Matlab program I don’t know how to put that kind of load (triangular load), and in the c++ aplication, I can put a partial triangular descending load, but not the ascending load.
Thanks a lot!
Sebastian

By: Dave

Dave — Sun, 23 Jan 2011 16:09:53 +0000

Dear Collin
Thanks a lot for that quick answer. You know, I’m just a, let’s say, «normal» engineer. I was looking around to find some Information about how to programm something like cba. (And found your page coming from sourceforge.net.) We’re planning a lot of excavations. A state of the art static-software ist normally like shooting with tanks on sparrows for the calculations we need.
So, I was just doing some checks to see, if your code would do, what I expected, and found the result above.

Thanks a lot, that was very helpful. I’m gonna try, if/how I can use it for my daily work.

Yours, dave

By: Admin

Admin — Thu, 20 Jan 2011 20:12:17 +0000

Dave,
Looking into this a bit further, CBA works fine when used as originally intended when it uses one ‘element’ per span. When you removed the central support, you made a single span with two elements. I agree though that this should be possible and so I have a fix that will enable this to be used:

1. in the file mbr_values.m, delete (or comment out) the lines:
d = d + disps(1).*((2/L^3).*x.^3 – (3/L^2).*x.^2 +1); % end displacements
d = d + disps(3).*((-2/L^3).*x.^3 + (3/L^2).*x.^2);

2. In their place, insert the following:
h = L/npts;
Ri = disps(2);
di = disps(1);
R(2:npts+2) = cumtrapz(M(2:npts+2))*h/(E*I) + Ri;
d(2:npts+2) = cumtrapz(R(2:npts+2))*h + di;

This is using the moment area method to solve for the displacements instead of displacement functions. It will give the same results as previous for single-element spans, but unlike the previous code, will work when more than one element is used to represent a span.

Hopefully you find it helpful.

By: Admin

Admin — Tue, 18 Jan 2011 16:17:11 +0000

Dave,
Very good spot – I’ve looked into it and the calculation for deflection is definately putting out something that’s not correct! The stiffness calculations are fine (e.g. the d of line 64, CBA.m), put in post processing for deflections along the member there is some error. It’s not jumping off the screen at me so I will have to take more time at it. I’ll post up a fix soon.

Thanks for that!

Colin

By: Dave

Dave — Sat, 15 Jan 2011 21:19:44 +0000

Hi
Thanks very much for this nice introduction to matlab and stiffness method. I enjoy it.

But… If I create a double-span beam, everything looks right. When I kill the middle support by setting
R = [-1,0,0,0,-1,0]
I’d assume to get a single-span beam. Moments ans Shear-Force are looking okay, but the deflection seems a bit … uh … strange.

Looking closely, there seems to be something wrong with the deflection near a support, even the ending supports: there is a change of direction, in which the curve is bent. (Excuse my technical english, please). I’m not sure, if this happens in a «normal single span», too. The change to see it on the deflection-curve is too small.

But, again. Thanks a lot.
~dave

By: wahyoe

wahyoe — Sun, 20 Jun 2010 07:37:49 +0000

how to use make source

By: Admin

Admin — Wed, 26 May 2010 12:46:05 +0000

Raim,
Once you have specified enough boundary conditions to the make the beam stable, it should work fine. If it is a support at only one end then it must be a fixed support (or at least have sufficient spring support stiffnesses to prevent large rigid body motions). Make sure both translation and rotation are prevented as a first step.
Colin

By: Raim

Raim — Wed, 26 May 2010 02:50:48 +0000

Hi, may I know if the CBA function can be used for a beam which has one support only at the end? I have found the function very useful for beams with support at either end, and kudos for that! But please enlighten me on this issue.

Thank you