Looking for engineers to discuss preparation of design calculations using EXCEL

For simple beam calcs I use a program called Beamchek

When the Excel drawing tools fall short, pasting Autocad in as a picture works well. Manipulates just like a picture, but makes big files. Prints like a charm.

-Rich

Mathcad by Mathsoft Inc.

Thanks for responding and links to the software. Unfortunately, this does not answer my question about EXCEL programming. I use EXCEL because I like to be in control of what is between "input" and "output". And I am not alone - I've just checked "excel group:sci.engr.civil"...

Check out MathLook for Excel by UTS for using 2D ("math like") notation. Also makes it much easier on the eyes when checking your formulas .

Their other product TK Solver is a powerful rule based multiple equation solver I used in school quite a bit..

Fred

Simon:

While I appreciate your desire to be in "control", the links to the software I gave you allow you to work efficiently with a minimum of wasted effort. Excel is a fine tool and I would be lying if I said I never use it for engineering work. However, for straight production design the two "beam" programs I gave you links for are very easy to learn, with intuitive data input and with easily understood results. I usually don't ask for more than that from a computer program.

Where possible, I try to avoid excel--i've found (to my dismay) that it's difficult to check for errors, and that once the errors get into the sheet they tend to propagate.

IMHO, this is problem is due to two things--

In recent years, I've been using MathCad--I find that it's much easier to see the "intent" of the calculations, and as such much easier to check for errors. Also, the printed output carries enough of the information that (if necessary) you can recreate the calculations from the printed output--so the printed documentation for the package is enough to adequately document the calculations.

I agree with another posting suggesting the use of Mathcad by Mathsoft. It is a little pricey, but you may get lucky and find an older version for sale on ebay.

It it extremely useful for the kind of repeat calculations you describe and you dont need complex formulas and relationships as in Excel. It also readily accepts drawings from Autocad or Imagineer.

You can easily set up your formulas,then copy and paste as needed, changing parameters easily. It is also very good at handling units and preventing related errors.

Good luck

Jose I. Villarreal, P.E. San Antonio, Texas

Excel certainly is useful, and I certainly use it every day for some stuff. But it isn't perfect. A few years ago I noticed that for some calculation we had in a contaminant transport model, that it was returning incorrect results. Turns out the ERFC function was not correctly implemented. Below is something I wrote internally back in 1999, talking about Excel 97. As we are still using Excel 97 here today, I haven't pursued this in any of the newer versions.

The bottom line, is if you do use these tools (or even paper and slide rule), you are still responsible for the outcome, so you had better be sure that your software is doing what it is supposed to be doing!

-----

For those of you who might be using the ERFC function in Excel 97. It does not provide the correct result when the argument is less than zero.

For example:

ERFC(-1) = 1.842700735 but EXCEL gives ERFC(-1) = -0.157299265

ERFC is defined as ERFC(x)=1-ERF(x), however when x

---------------------------------------------------------------------------- Aug.28,2003 Answers to thread (messages from 4 to 10): Looking for engineers to discuss preparation of design calculations using Excel

Google ? Groups:sci.engr.mech, sci.engr.civil

-------------------------------------------------------------------------------- To: Bob Morrison (mess.4)

Bob: Sorry for terse remarks I made previously. My points are:

Other points which attracts me to Excel are that it is very flexible, popular (anyone could use a program I did), very portable (I could come to any office and start working), allows for programming any problem, has sketching capabilities and emulates manual design. It has many drawbacks, of course, which I am trying solve (one approach is to know opinions and suggestions of others)

Simon

-------------------------------------------------------------------------------- To Rich Jones (mess.5):

Rich:

I'm trying to simplify the process and make it fast. And I am talking is not about a CAD drawing but a simple design sketch you are doing when making it with pencil on paper. When a CAD drawing has been done it's probably already late to prepare a sketch for it. Also not everybody has AutoCAD, or ability to work with it. What about portability? Can I send it to my friend if he does not have AutoCAD

Simon

-------------------------------------------------------------------------------- To: Paul Ranson (mess.6)

Paul:

I am talking about Excel. Maybe because I am in love with it? We can talk about Mathcad, Mathlab, Mathematica and others next time. My initial question was: how to type a formula (in EXCEL) in a usual math format and not re-type in the formula window. Such Add-ins for EXCEL as MathType, Equation Builder, MathLook and others are typesetting programs and have nothing to do with formulas. TK Solver is great also?

Simon

----------------------------------------------------------------------------- To: Fred Ernst (mess.7)

Fred:

My initial question was: how to type a formula (in EXCEL) in a usual math format and not re-type in the formula window. Such Add-ins for EXCEL as MathType, Equation Builder, MathLook and others are typesetting programs and have nothing to do with formulas. TK Solver is great also? Simon

------------------------------------------------------------------------------ To: Michael (mess.8)

Michael:

I absolutely agree with your criticism of Excel. I started this discussion to find opinions and suggestions to improve it. I don't know yet how to achieve it. One of the ways is to make formulas readable in cells as math equations (not as a typesetting pictures). One of the typesetting add-ins to Excel (Equation Builder) is very close to it. There are placeholders (similar to text boxes) where you place numbers or letters. Excel's Text Boxes allow for connections to other cells. I don't know? maybe it is time to develop another version of Excel targeted to engineers? And I know MathCad ? it is great, I have it, but I have found that not many of my colleagues have it or wishing to buy or learn it? It would be not as popular as Excel

Simon

-----------------------------------------------------------------------------

To: Jose I. Villarreal (mess.9)

Jose: Please see my response to Michael (message 8)

Regards, Simon

------------------------------------------------------------------------------

To: Nickolas Fitzpatrick (mess.10)

I agree that Excel is not perfect. It was developed for business/financial? I see a lot of changes to be done to better serve engineers? But nothing is perfect, anyway. Very often, especially if the result is suspicious, I make a fast manual cursory checking Maybe, we will find a way to approach Microsoft to develop engineering (scientific) version of Excel

Thanks for a note on ERFC Simon

------------------------------------------------------------------------------ Thank you all

I believe that is exactly his point, do you trust a high price program just because its "high price" ? If your going to stamp and put your name on a design shouldn't you be the one doing the math? How do you trust a program if you don't do the programming? Where's the line between cost-effectiveness and dangerous.

Just wandering;-) rentapen

.>usually don't ask for more than that from a computer program.

Dear Rentapen:

I hand full of free advice:

- you get what you pay for

- if you are paid to develop software, fine. If you are not, they'll find someone to do your regular job.

- if you have a software "suite" that does the job, validate that you get results for cases you can test. then lock the suite down.

- if it works don't fix it.

Home projects are well and good. It feels great to develop your own algorithms. But don't limit your company to only your efforts. Your competition will not be afraid to use predeveloped tools. Move fast, move quietly, and bury them. Business is not always about "feeling good". It is about being as right as you are allowed time for.

David A. Smith

Both the programs I recommended have results that are easily checkable with hand methods. We are not talking about complex FEM programs here, but simple beam and column programs. I guarantee you that I can outcalc you using these programs --- and in this business speed can equal time which can equal money. I have written some complex spreadsheet routines to calculate lateral forces on residential structures. I spent a lot of and effort time on them and when a new building code came out they were no longer any good and had to be rewritten.

If you do a lot of concrete column design then by all means buy a program that does that type of design. My point here is to spend less time writing programs and more time doing design. You don't seem to want to trust the software writers (I might add that is a healthy skepticism). However, If I was your boss I would be less willing to trust complex spreadsheet programs that have not been fully beta tested by someone who didn't write it.

Sure these are all powerful tools and they have their place in the scheme of things. But how many times during the day are you designing a simple beam versus the number of times you are designing complex structures. If it is the latter then the programs I recommended are no good to you. And I am quite sure that it would be a major programming effort to get Excel or any other spreadsheet to even come close to what the FEM programs will do.

As another poster has said "- if you are paid to develop software, fine. If you are not, they'll find someone to do your regular job." As I alluded to above, if you worked for me your job would be as a design engineer not a software writer. If you insisted on spending enormous amounts of time on writing spreadsheets you would soon no longer be working for my company. So, you have to decide if your job is a software writer or is it a design engineer.

You do it the same way as if you were doing the programming yourself..check the results of the code against known results..preferably analytical but if none are available then against published results.

------------------------------------------------------------------------------ To: Rentapen (mess.11) and Peter Attar (mess.13): "Check the results of the code against known results.. preferably analytical but if none are available then against published results". (Peter Attar, mess.13)

I don't think it is practical and enforceable in an engineering office. Managers would consider it not a "regular job" for engineers. At most they will pay lip service to it. And this can not substitute for proper checking anyway.

-------------------------------------------------------------------------------- "If your going to stamp and put your name on a design shouldn't you be the one doing the math?" (Pentapen, mess 11)

This is not a simple question. Are we going to reject the whole industry of commercial programming?

My suggestions:

Simon Goykhman, P.Eng.