Descus vs MDX 2

baker12, I have used MDX for about 10 years and up to this time it was very good. Also, the expertise of Mr. Richard Douty was excellent in helping with advice on using this program. However, his son Chris has taken the existing program which was excellent and tried to improve it. In the process he has managed to produce a program that is more difficult to use and is not backward compatible with the previous versions. Also, I have modelled some of our curved girder bridges on STAADPRO to compare with their new grid system program and the results were not good. I am looking at going with DESCUS I at this point. Ed McFadden

Well, there's a problem...using staadpro to check curved girder analysis. In my opinion you're putting more faith into that program than I would.

I have used both programs and have found MDX to be more user friendly with input and output. Additionally, I have had good comparisons with both programs against each other and with GTSTRUDL.

I should have also noted that the support with MDX is really good.

Qshake

baker12, If DESCUS I is more difficult to use than the Windows version of MDX, then I will just use STAADPRO to design my curved girders. Actually, we have used STAADPRO to model two curved girder bridges that are in service. The actual deflections are very close to the models. I guess in depends on how one uses a program to model the real situation.

baker12 and Qshake, I could not figure what was the difference between the dos version of MDX and the Windows version. I ran the dos version and at the pier I got a dead load moment of 372.55 K-FT; the Windows version got 708.62 K-FT. So I modelled the bridge on STAADPRO and got 373.5 K-FT. Now, which of these looks like the correct answer? I think the new version of MDX needs a little work.

I have received many comments on the output of Windows MDX vs DOS MDX. Everyone seems to think that there are still some bugs in WIN MDX. We currently use WINDASH for I Girders and since the user interface is very similar, it looks like we will go with DESCUS II for our curved box girders.

Gentlemen:
I have just finished a month long evaluation of a bridge that was designed with an older version of MDX. I used Staadpro 2003, MDX Version 5.98, and current Descus I. Staadpro 2003 got the best results, then MDX, and last came Descus. I used the high resolution finite element option on MDX and just used Descus which ,I think, does a grid analysis only. The main differences occurred in the Live Load output of the programs. One item that bothered me about Descus I was that lane loading produced a higher than expected girder stress at midspan of span 2 which was only 79 feet long. I modelled the lane load on Staadpro and got only about .5 of the moment due to the truck. MDX has worked with me to get the newer version 6 debugged; so I am staying with MDX for steel girder design.
Ed McFadden, P.E.

baker12:

I wouldn't buy any of them.
MDX graphic interface is really terrible. It use simple rod elements but they call them "finite elements". For programs like this, Aeronautical Engineers call us “low tech engineers”.
Stick to a real engineering program: SAP2000 or GTStrudl.

CMFG

Gentlemen,
Thanks for everyone's input. I should have been more specific in my original post. I was looking for input of MDX vs Descus II for curved box girders. I had used BSDI in the past and was investigating other programs. After using Descus II on a trial version, I have concluded it is not for us. I would pick BSDI over Descus II anyday.