Structural Steel Design Software - Opinions? 1

[cal91]

I'm relatively new to my firm, hosting a total of 3 engineers including myself. The way they have "always done things" is the one has Sap2000 on his computer, and the other has Revit. Mr. Sap2000 creates a model on SAP (from scratch) and does the analysis/design. Mr. Revit creates a model on Revit (from scratch) to create the drawings and the coordination model.

We have a new project where we are building a steel building on top of a concrete building. We're the steel engineer and another firm is the concrete engineer, and we will need to coordinate models. They asked what program we used and I told them Sap2000. That got some chuckles to my surprise. They asked if we used RAM or ETabs, which they much preferred, and which we don't have.

This got me thinking, especially since Mr. Sap2000 is retiring soon and I will be taking his place, is Sap2000 the best program for us? We do heavy seismic steel design, and I'd like to be able to just create one model, either in Revit or the Analysis program, and just share it between the two softwares. I've looked at previous threads on this but can't find anything since 2011, which was centuries ago as far as technology is concerned. Also, why did they chuckle. Whats wrong with Sap?

 

[TEDstruc]

I design mostly non-building structures such as heavy industrial type work. Risa 3D works great for this type of design.

Currently our work flow goes something like this:

1. Steel is laid out in 2D using autocad
2. 2D steel layouts (.dxf files) are imported into Risa 3D as a drawing grid. Steel members are graphically drawn in Risa 3D using the drawing grid imported from auto cad as a template. Alternatively, you could directly import the .dxf which will automatically convert all lines to members, but you would then have to modify each member for appropriate end conditions, material, shape, etc... So I prefer to directly input every member so I know everything is done right. Risa 3D model is used to size or verify member sizes and make improvements to layout.
3. The 3D model is also created using the 2D autocad file as a template, we use inventor for this because we do our own fabrication detailing. We have also dabbled with advance steel for this, but still haven't managed to make it efficient enough to fully convert.
4. Installation drawings are dimensioned and annotated in autocad using the 2D layouts as xrefs.

Generally we feel like there has to be a more efficient work flow, but up to this point we haven't found it.

Anyone else doing there own fabrication detailing? What does your workflow look like?

 

[TehMightyEngineer]

IES Visual Analysis

I took a harder look at IES Visual Analysis, seems fairly comparable to RISA, maybe slightly less features in the default package. I agree that if you're sharing licenses between parties then a subscription or other network license makes much more sense. It also appears from what you say and from their page that their subscription license package is reasonably priced and probably worth it over the stand-alone.

Definitely like that if you pick a bunch of advanced options and features you don't end up spending $6,000+ per month (as RISA seems to now be priced at if you get the floor and foundation features).

Maybe I just need to bite the bullet and accept that subscription licenses are the way the industry is going. It's a shame, having options is always preferable to no options. If they don't like the stand-alone profit margins then raise the stand-alone prices, but eliminating it entirely is doing their customers wrong in my mind.

While I like many things about RISA, one thing that I've not particularly enjoyed is the interface.

Workflow often feels non-intuitive to me and it almost feels as though it was programmed in Windows 95 and never got fully updated.

Clearly you've never used STAAD. That program is a dinosaur that's been beaten into the modern age. RISA is smooth and intuitive compared to STAAD.

That said, from looking at the promotional material for IES's software I would agree. Their layout and workflow appears slightly more intuitive to RISA. Though, for me only STAAD's interface was bad enough to be a dealbreaker. I can suffer a lot if the features or benefits are there.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[JoshPlumSE]

For non subscription options for Structural software, I believe the only options are:
1) A currently owned version of RISA, though if you ever let your maintenance lapse you will be forced to move to subscription. One of the pleasant changes brought to you by the new management.
2) SAP / ETABs - I believe Computer Structures Inc (CSI) still allows you to buy a hardware key for a stand-alone version of the software. Though I am not 100% certain of this.
3) Visual Analysis. I'm sure the new policies over at RISA/Nemetschek and RAM/STAAD/Bentley have really helped sales for the Visual Analysis folks.

I don't know of any others for certain.
4) Robot is owned by AutoDesk, so I assume it's all subscription.
5) SCIA is owned by Nemetschek, so I'm assuming it's also subscription only.
6) Dlubal software (US office in Philadelphia) I believe still offers stand-alone licenses. Though you'd have to ask Amy Helig their US sales rep.
7) Midas is popular in Australia I believe. But, I don't know what kind of licensing they have.

 

[KootK]

and probably worth it over the stand-alone.

No actually. If you do the math on it, stand alone is still much cheaper over the long haul because the annual maintenance fees are very low. IES is clearly set up for perpetual license as the dominant model. I'm just a commitment-phoebe is all. And to clarify, you can share both subscription and perpetual ownership licenses among parties. At least I can. Every man for himself.

as RISA seems to now be priced at if you get the floor and foundation features

That is the kicker as RISA only work for me if I also have FLOOR. VA is not as feature rich as RISA but it does have base functionality that covers most of what I need from floor. I just want to have easy, area loading style load takedowns throughout modest sized buildings. I just don't have the patience for tedious load take-downs any more. If I can't automate it, I'll make it pretty damn rough.

 

[Agent666]

Spacegass is another option without a subscription model, maintenance is only equivalent to 350USD a year and that includes all upgrades to new major versions (which some other companies charge through the nose for, I'm looking at you CSI ...). Its essentially a perpetual license and it's up to you if you pay yhe maintenance fee.

It's fairly popular in Australia/New Zealand, especially in replacing Bentley offerings.

 

[Celt83]

...I just don't have the patience for tedious load take-downs any more. If I can't automate it, I'll make it pretty damn rough.

Koot if you've got Revit give Dynamo a look I've created some simple scripts to do load takedown for our wood jobs, requires some custom wall properties and input for trib widths from the users but after that can do the load analysis and wall stack designs for a fairly decent size 5 story wood job in about 30 seconds (1 day +/- taking into account the manual input for the tribs)

Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

 

[271828]

"2) SAP / ETABs - I believe Computer Structures Inc (CSI) still allows you to buy a hardware key for a stand-alone version of the software. Though I am not 100% certain of this. "

I am pretty sure this is correct. However, the maintenance agreement is $875/year. It's required for updates.

 

[TehMightyEngineer]

A currently owned version of RISA, though if you ever let your maintenance lapse you will be forced to move to subscription. One of the pleasant changes brought to you by the new management.

I'm kicking myself now that I didn't get maintenance for RISA. My plan was to buy RISA and use it until I needed something more advanced like FLOOR. Then purchase floor and upgrade my license to include maintenance when it made sense. Essentially an extended trial. Probably can't do that now from what I've heard and not sure I'd want to anyway.

If you do the math on it, stand alone is still much cheaper over the long haul because the annual maintenance fees are very low.

Yeah, I meant worth it in non-monetary considerations. You spend a bit more but can switch to RISA or SAP or whatnot if your needs change and the initial setup costs are lower which is useful for a startup firm.

Ian Riley, PE, SE
Professional Engineer (ME, NH, MA) Structural Engineer (IL)
American Concrete Industries

https://www.facebook.com/AmericanConcrete/

 

[MrHershey]

2) SAP / ETABs - I believe Computer Structures Inc (CSI) still allows you to buy a hardware key for a stand-alone version of the software. Though I am not 100% certain of this.

Don't know on standalone but for network it's not subscription. You can let support lapse and keep asking them for new network license codes when they expire or you switch servers for as long as you want I think. You just won't get updates or support (other than the new license codes).

 

[KootK]

Koot if you've got Revit give Dynamo a look I've created some simple scripts to do load takedown for our wood jobs

I'll check it out Celt, that actually sounds pretty great. I would love for my load takedown stuff to be FEM software independent. For better or worse, it's not as though I'm ever going to not have a Revit subscription.
Are you able to tell me any more about it? I want one of your models quite desperately but I don't imagine that's going to happen.

- Can you do beams and columns this way too?

- Can wall loads be transferred onto beams?

- Call wall loads transfer laterally at floor level offsets?

My thoughts on this aren't yet fully formed but I strongly suspect that the future of FEM software will need to be this:

- Simpler but crazy powerful FEM software that does NOT do element design but does a great job of exposing the internal databases to the user for export & import. CSI is pretty good at this really.

- External tools, both commercially available and user developed, that do the member design outside of the FEM package, passing data back and forth.

Since the beginning, the vision has always been "all of your design needs under one roof". That's clearly a mistake in my opinion and I think that the struggles that software vendors have had trying to be everything to everybody are solid evidence of that. I don't think that any FEM vendor is ever going to be able to give everybody the flexibility to do what they want, how they want it. At the same time, contemporary design engineers have never been better at programming etc.

At the risk of striking terror into some hearts, I expect that the FEM side of it will ultimately be open source and/or freely available somehow. If it's not done by the time that I retire, I might just up and do it myself.

 

[Celt83]

the short answer is Dynamo is a visual programming way to use and manipulate certain aspects of the revit api. IE need to know the start and end coordinates of a beam fire up dynamo add 2 block pick your beam and hit run and you have the x,y,z coordinate of the start and end of said beam can repeat the process do grab the analytical model.

The dynamo stuff has all been on company time with company software so probably a slim chance I can do more that describe the process.

The hitch with FEM is stiffness all our hand methods for bearing wall load tracking is simple tributary width x surface pressure with some simple beam theory thrown in for line or point loads within the span. Along with wall stiffness is the slab stiffness what happens with a diagonal section of building you've got to adjust the x,y shell stiffness to get the distribution you expect. So on the Dynamo end my aim is replicate our hand process I add parameters to wall that let the end user tell it how much tributary width it supports at each floor, another parameter sets the floors uniform loading, and another parameter for any miscellaneous loading (this captures those point loads, etc.) the script then reads all this in performs the per floor load take down and writes the load per floor and total base loads back out to each wall to be scheduled. For wood buildings along the way it uses the logic from my wood stud wall program to design stud spacings at each floor and write that to the wall this can be scheduled and if you uncheck "itemize every instance" in a Revit schedule you'll have the reduced wall stack schedule for the whole project. It doesn't fully automate the process but we've noticed a pretty decent decrease in load take down time.

For podium projects there is another layer to the dynamo scripts that writes the load data using the wall x,y coordinates to a staad gcf file which can be directly imported into concept so no more tedious manual application of line loads for transfer slabs.

to address your hyphens:
-Beams and Columns - Haven't done it but yes, in a two-way slab system can get even more automated with a monte-carlo algorithm to get pretty good approximations of tributary area automatically and by extension handle live load reduction.

-Wall loads transferred onto beams - in theory possible but I haven't attempted it yet

-Wall loads transferred laterally at floor levels - again in theory possible but haven't attempted it yet. would need to know floor span direction and programmatically pick the correct supporting walls.

Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

 

[sticksandtriangles]

Celt83,

You've got me excited with what you've done in dynamo. Great IDEAS!!!

I actually started playing with dynamo on the side and on the company dime and have yet been able to sell them on the possibilities that it has (I've created a concrete column schedule through dynamo and concrete beam schedule dynamo so far, but even I will admit this is really rough and one has to truly know what they are doing in the dynamo script for it to work). The ideas you have presented seem like some powerful capabilities that may get them to bite on the possibilities.

I might try to dig into this based on the little that you've provided so far and see what I can do.

It's crazy that my coding hobby can (in my opinion) can make someone stand out in a structural engineering profession.

 

[Celt83]

sticksandtriangles:
FYI you can use Revit and Dynamo to their full capability with Revit in viewer mode, so if you feel like poking around at home just grab the latest of both and install the demo version from Autodesk.

Find some repetitive task you do often in Revit and use that as a stepping stone, my first pass was drawing 10'x10' drop panels at column locations with a top offset equal to the slab thickness sounds simple but has several gotcha moments with how revit defines columns and slabs.

Code Blocks will fill some voids and make things much easier to follow sometimes.

For the really advanced stuff your looking at using a python block and accessing the Revit API calls directly.

Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

 

[sticksandtriangles]

I was hoping to pick off an easy one as well Celt, the concrete column schedule, but it became a beast of a dynamo script.

I had to get into python code and access the Revit API as you had mentioned. Definitely not the easiest one to target. Presenting this to other engineers definitely "intimated" the user,drafters and myself honestly. End code had this appearance so i can understand where they were coming from haha:

 

[pmtrevisan1]

fellow engineers
I was in the same boat as many of you.
when i finally went on my own last year risa was subscription based. bit the bullet and got the 1 yr license. though I like risa I'm not a fan of any subscription based software.
I finally ended up buying s-frame.
so far quite impressed. they just released s- timber that does wood design.

only complaint thus far is lack of a proper manual or tutorials. risa is awesome at this part.

s-frame handles all the analysis of all the other softwares and is purchased and not subscription based. and ties their modules together for fdn steel concrete etc.

their customer support is absolutely amazing.
if anyone is in need of an owned software have a serious look at s-frame.

cheers.

p

 

[Celt83]

Ha...yeah that looks about right. I've found schedules tend to work out better using the Revit schedules and adding additional shared parameters to families as needed, for everything except concrete columns have yet to find a good way to schedule those in Revit.

It was a hard sell on my wall tracking stuff but what I did was group anything that didn't require user input into one big red box and labeled it "magic don't touch" and put the 4 or 5 actual user inputs in a green or blue box and pushed them pretty far apart so when it opens you have to zoom and pan to get to the business and that got a chuckle in the staff meeting and then a "Oh that's all we have to do"

I've got another one that reads concept bar output and places an instance of our top or bottom bar family at the correct location with concept designed length, bar size, count, and hooks then halftones everything as a method to indicate the bar set hasn't been reviewed yet. In theory the Design engineer then reviews the plan and removes the halftone to indicate its been reviewed making edits and combining middle strip bars as they go. Limits the user error of missing a bar set here and there or copying the wrong info from another floor and a little piece of mind that if not reviewed then at a minimum we're getting the bars that our design programs expects to be there.

Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

 

[sticksandtriangles]

Interesting, you make the engineer open dynamo?

I was trying to limit anyone opening dynamo and just having the drafter or engineer hit run on the dynamo script.

Also feel free to ignore this part, but I am also developing something that reads concept bar output and places revit families in the same location on a sheet (bars size length etc).
Do you send the concept output to a cad dwg and read the x and y coordinates of this element?

That is what I am currently exploring and it looks promising.

 

[Agent666]

I've used Dynamo as well and believe it has a huge potential to take repetitive tasks and automate them.

A few years ago I created a dynamo script that sequentially renumbered and tagged about 300 piles with pile type/size etc on a project and reported back the nearest grid intersection in a tabular format to include on pile schedules. Saved a huge amount of time when compared to doing this manually and was basically error proof. A few dynamo revisions later and it was broken and I never went back to fix it.

Other things I did was try convince our residential waffle slab designers that creating the plan drawings was as simple as importing the waffle slab outline as point/line data and hitting a button and program draw the slab, the ribs at whatever centres each way, perimeter beams, etc. But they were too used to tracing stuff in autocad to see the potential for automation.

But some of the draftspeople where I last worked were into it big time as well, they had whole processes for uprevving drawings and creating schedules of drawings for a drawing transmittal from dynamo. Most developed in their own time. Saved a lot of time vs someone manually having to type this sort of stuff by hand, its just a really good way of interacting with the databases that lie behind Revit. Unfortunately it all seemed too hard for a lot of the old school people to operate, and old school management really didn't get it, especially the types who didn't think much of the Revit workflow in the first place.

I think it has great potential, I'm imagining you can embody for example design calculations within dynamo to be applied to a selected beam, that sizes the beam and replaces it in the model at the click of a button. The potential for giving someone a layout and getting back a rough size of beams from the draftsperson as a starting point is really good. I'm sure if you put in enough effort you'd be able to create scripts to find the midpoint of a beam, work out the nearest member either side orthogonal to the midpoint and based on load, this tributary width and beam length work out a size/deflection etc and update the size. I remember watching a video years ago of some complex structure that used dynamo to manage the structural optimisation (possibly linking to Autodesks Robot analysis software, can't remember exactly).

I found the documentation was quite hard to understand for some nodes, basically the syntax of what inputs and outputs were required took a lot of trial and error to get the result you required. Often in hindsight you were like, 'oh thats what they meant' or 'geez how the hell did I ever figure that out' or 'I simply give up, what the hell is going on'. Often you had to solve your particular problem in some round about way, when you wer elike why the hell isn't there a node that does this. Mind you this was all on 0.9ish versions, things may have improved a bit since then.

I believ eyou can also incorporate nodes that include your own python code which opens up the possibilities for customisation.

If you go online and check out some of the Autodesk University video presentations on dynamo it will give you a bit of a flavour as to what it can achieve, its not all about generating complex geometry (though its used/good at this as well).

 

[Celt83]

Sticks:
export the reinforcement tables for Latitude and Longitude concentrated program bars. I reduce the table headings to just "Number", "Location", "Bars", "Bar Size", "face" and "Length".
You'll essentially get a csv file from that and the biggest pain is the location since it gives you 6 coordinate sets I forget off the top of my head which one is actually the bar center point but it's thankfully consistently that coordinate set for all the bars.

Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

 

[sticksandtriangles]

You're awesome celt! Thanks so much.

I did not even think about that process, saves the go between of reading a .dwg file.