About BSDI

[kji1204]

I spent some time searching for any information on BSDI, but I couldn't find any.

Anyone who have info about BSDI(catalog, web site, or anything else),plase give some advice.

Thanks.

 

[LCruiser]

What is BSDI supposed to be?

 

[cooperDBM]

Google BSDI-3D or Bridge Software Development International. They're based in Pennsylvania.

 

[Umassengr]

BSDI is more of a service than a program. As far as I know they have not developed commercial package for this software. Each job is customized for the client by BSDI.

You will need to work with BSDI to develop the model of your bridge in their software. Basically you send BSDI the member properties at relevant sections and a layout of your structure model along with some pre-processor data BSDI will develop the 3-D model, run it and send you results. You will then refine your model though subsequent iterations until you get the results you need. It is generally two to three iterations.

I assume with the advent of highspeed internet connections and increased capacity for sending large data files over the web that this can all be accomplished via email and FTP sites. We used to send 3.5" floppy disks back and forth so the time lag was quite long.

A note:
I last worked with these guys about 9 years ago to develop models and designs of highly curved trapezoidal steel box girders. BSDI is very good at what they do and the software gives very good results, but be sure that this is what you need for your project because as I describe above you do not just order the software, open the box and go to work. Good luck.

 

[Qshake]

Umassengr - How do you know if the software gives good results?!

And more importantly if you say that you've checked it with some appropriate general purpose FEA program how does the little firm with no software check the work or guage what is given to them?

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[Umassengr]

Qshake
The point I was trying to make in my original reply was that kji120 make sure that the services that BSDI offers were really necessary.
As I said this is not an off the shelf software, it is a company that provides a service. They are essentially part of your project team. So if you are hiring them to help you, you have to put a certain amount of trust in their results.
When you work with them it feels like you are working with a black box program, but then in some ways so is every other off the shelf program that does curved steel design (MDX, DESCUS) and I know designers who have found errors in the results of both sofware packages and have had to call the software support team to help them out. The difference with BSDI is that you are already working with them to get results customized to your project. We used BSDI to help us with our project because we could not get the results we were looking for with a normal FEA program, or commercially available software. I am sure given enough time and money we could have done it, but we had limted amounts of each.

 

[Qshake]

Umassengr - I agree with your last post, in that it would feel like a blackbox program. I just wanted to know how all this worked since I haven't worked with them before and am not knowledgeable of how quality control is assured. There seems to be a lot of venues for errors...say in just describing the bridge. I assume they don't come to your office.

Many contracts with DOT's note specifically that the general engineering consultant is responsible for the quality of the subs. So again, I was simply curious how one is able to establish this beyond 'trust'.... Trust is ok, until you get sued.

Thanks for your input.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[graybeach]

I have worked with BSDI on a couple of curved, skewed, I girder bridges. When you use BSDI, you really have to understand your structure. It is practically impossible just to spit the results out onto a drawings. First you have to input the geometry using their convoluted conventions - not easy. Then the loads - also not easy. Once the analysis is done, you get this dos-like post processor that is extremely obtuse to use. You end up having to do a fair amount of approximate hand calculations just to understand how the sign conventions work. Then you have to really wrassle with the thing to figure out the camber. It is amazing how the irregularities pop up when you have to delve into results like this. Once I went through the whole process, I felt very confident about the design.
By the way - when you go to their office, you will have lunch at Dann's daughter's restaurant. Stay away from the quiche.

 

[Umassengr]

Graybeach's post illustrates the process better than my memory could recall. I had the cheeseburger if I remember right.

Qshake -I think the question of trust that you raise is applicable to many of the industry "standard" software that we are all using.
The design software that my company uses was all purchased by previous department managers there and it is trusted to deliver correct results, so we trust it by experience. I do not perform second sets of calcs or independent parallel designs to verify the software, other than to check the input and/or when it gives unexpected results.


With a company like BSDI the trust comes your interaction with the BSDI staff in making the design work.

 

[Qshake]

Umassengr - Most conventional software programs have text files that are checked at my company for accurate input and that is performed on a project basis. We do actually benchmark all software once when purchasing it to ensure we get results similar to other programs we have and some that we've programmed ourselves.

When reviewing our data from programs on a project basis we check the output against rules of thumb and back of the envelope type sanity checks.

We do not accept that the results are correct because Joe on the 4th floor did the file and he's a whiz with things like this nor more than we accept subconsultants work with out sanity checks. The latter is similar to what BSDI sounds like.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[Umassengr]

Qshake
We do all of the same things. I was not suggesting that just because a computer program or a subconsultant says the design is right means it is. Would you agree that the reputation of the software goes a long way in your having confidence that the results are accurate? I have used software out of the box based on its reputation. I've done an evaluation first but the real vetting of the software was done in version 2.0 and then the industry accepts it as must have software for design. My point is once the validity of a program is established, other than verifying program input and doing sanity checks of the output, how far should you go to verify the program mechanics? In the case of BSDI: If the industry considers their services and reputation as acceptable, outside of the input and output checks, how much more QA is needed?

 

[Qshake]

Umassengr - I see your point. It is unconventional to me and that's why it strikes me the way it does. Plus Graybeach's comments on the unconventional nature of the output.

That said, once we've rolled up our sleeves it's back to input/output(sanity checks) and we're done. Black box remains until someone finds a problem it doesn't do so well (or so we think...can be operator error!). Which is the same as BSDI, give the problem to them get it back and check the input/output call it a day.

Regards,
Qshake

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[Umassengr]

Exactly.