Electronic calculation production 9

[Agent666]

Hi all.

In the last couple of years I've seen an increasing number of businesses moving towards embracing producing calculations by electronic means. By this I mean rather than using pen and paper and handwritten calculations and scanning them in from years gone by, the calculations are for the most part typed out on a computer with screenshots taken from analysis programs or drawings. Alternatively with the advent of tablets and computers with touch screens someone can write direct to the screen.

I've always thought you lose something regarding understanding the thought process of the author of the calculations when calculations are typed out because it seems to take longer to type out formulas and mathematical calculations. So the calculations become more brief as people are pressed for time, and ultimately the calculations become harder to follow and understand based on my personal experiences doing peer reviews of these types of calculations.

Basically I've resisted making this electronic transition so far, but time has caught up with me it seems. Basically moving into a team that creates all calculations electronically and the expectation that I do the same. They primarily use Microsoft OneNote for this purpose. After the calculations for a design are finished they are printed to PDF and a header/footer added to the PDF with company banner and page/project numbers, etc and sent off to local authorities/clients etc.

So does anyone else have any advice or do all their structural calculations electronically that can give their take on it or offer up their workflow suggestions to ease my pain. I've been given a Surface Pro and surface pen, and to be honest I am actually quite liking writing on the screen, I'm just not sure OneNote is the best tool for storing absolutely everything to do with formal calculations.

So far I've noted the following in no particular order:-

1) handwriting on the screen is actually ok, ability to go back and easily amend previous calculations is fairly handy (like correcting mistakes or inserting an intermediate page or paragraph within a page without having to write it all out again.

2) I feel you only produce a 'final' calculation when typing out the calculation, you lose the design development part and flow through the calculation, like try this beam, fails for some reason, then try this size and show it works. I like seeing how someone arrived at the final decision.

3) As I understand it the OneNote 2016 app is being retired in about a year in favour of the Windows 10 version which seems to have far less features. The others in the team seem oblivious to this, but I see it as a major roadblock with the workflow as they are quite different programs.

4) OneNote is primarily a note taking application, needing to jump through hoops to get for example a printout of a spreadsheet in there (if its multi-page spreadsheet then it gets even harder (see next point) seems really messy). You're basically forced to almost take a screenshot of Excel which is pretty messy solution, if you have a low resolution screen they invariably end up looking like garbage comparable to a 12 year old school project. I feel for the most part it was never designed to print out a PDF to scale from a OneNote page. Half the time despite writing within the margins it seems to cutoff text and throw it on a section page. Seems like OneNote with the default infinite canvas is all good and someone thought as an afterthought 'hey lets throw in some templates for different paper sizes' without really making it work seamlessly.

5) OneNote only allows a single page of text per 'OneNote page', no multiple pages of text in one 'OneNote page' if you want to print it out nicely at the end.

6) So far I've been mostly writing on the screen trying to follow my previous workflows rather than typing everything out. Most of my colleagues seem to be typing everything out as far as I can tell. This seems really inefficient to me, and because of the complexity of writing out math equations like this people seem to only write the final answer so you miss out on all the intermediate working which seems a recipe for introduction of calculation errors.

7) putting screenshots of drawings or code provisions and marking them up in the calculations is really easy, beats busting out my scissors and tape.

8) Tried using bluebeam, but the pen support seems to be lacking, pen just seems clunky/slow and sometimes causes you handwriting to glitch out at which point you have to press undo like 10 times just to delete the last word you wrote.

9) Importing a PDF printout into OneNote to write over is dead easy, however there seems no way to print it out again to an updated PDF with any markups given the infinite canvas of OneNote. Doesn't work with the one page rule above for example.

10) is there an app out there that is better suited to calculations than OneNote?

Anyway interested in your thoughts/experiences/suggestions.

Thanks

 

[BridgeSmith]

I've used Excel almost exclusively for my design calcs, as do most of the engineers in my office. We've found it has adequate options to import the output from the analysis software we use, and the functionality to do most of our layout and structural design calculations.

I've been printing my excel sheets, along with my hand sketches and any hand calcs, and then scanning the completed design. However, a colleague informed me that if I use the print-to-pdf option in Excel, and then assemble it into a pdf with the scanned sketches, etc. (requires the full Adobe Acrobat program), the Excel portion is searchable, making it more useful to someone referencing it later. I tried it, and had a few issues with the page sizes, but I'll get some help getting that fixed and start doing that from now on. He also suggested scanning the sketches and inserting them into Excel as images, and then doing the print-to-pdf, so as to avoid the need to assemble the pdfs into one file.

For some things, hand calcs are faster the first time, but if you're going to have to do those calcs again for another design, I've found it's worthwhile to take the extra time to do the calcs in Excel.

I've learned through much frustration that it's far better to keep input values in their own cells, and reference them in the formulas. Pretty much the only numbers hard-coded in my spreadsheets calculations now are constants in the equations out of the design spec.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[Celt83]

Since you have Onenote 2016 do this:
Create a page template one of the options should be page size when doing this in Onenote 2016

if your company has a calc pad scan in a sheet and apply it as a page background.

Then set that template as the default for a notebook/section.

Onenote 2016 is still going to get security updates for awhile, will edit this with a link to their post on the issue. While Onenote on Windows 10 seems simplified and missing features there are a couple that it does bring to the table such as the new math tool, you can type or hand right a symbolic or real equation and the tool will solve the equation and insert the solution. There is a ruler tool and syncing is faster and more reliable.


Other options besides Onenote:
Word - uses the same underlying inking system and gives you pagination. Lose the "notebook" feel.

Drawboard - pdf editor comes pre-installed on the surface. Decent pen support and can create blank pages as needed. Lose the "notebook" feel and any ability to share file and live co-editing.

Autodesk Sketchbook - app made for drawing/sketching but comes with excellent tools like a ruler, french curve set, perspective, etc. With the ability to make custom brushes you can easily draw repeating elements like cmu. My workflow here is bring in a scanned copy of a calc pad sheet and make a new layer for each calculation page.

Open Source Structural Applications:

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

 

[IDS]

I didn't know (or maybe had forgotten) that OneNote has a math solver app. A quick search suggests that it is also available for Word, but not Excel, which is a pretty weird decision even by Microsoft standards.

To my thinking the app out there that is better suited to calculations than OneNote is called Excel, even without the math solver app.

Update: A slightly longer search suggests that the app for Word only works up to Office 2013. It looks like in Office 365 it is only available for OneNote.


Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[Agent666]

Most of my work has traditionally been specific hand calculations with spreadsheets for specific element design (e.g. beam/column/wall design) mixed in with printouts from analysis software, etc. If outside of the scope of the spreadsheet some hand calculations usually suffice. I guess very similar workflow to what BridgeSmith noted. Even if I'm able to do it all with spreadsheets I still like to have some handwritten pages explaining what I'm calculating dispersed among the sometimes meaningless screeds of spreadsheet output.

All calcs scanned at the end and assembled into a single or multiple PDF's depending on size of the volume of calculations. Part of the reason I always liked having a hard-copy was I could have several pages spread out in front of me to reference back to previous calculations (Onenote and Word can also have multiple windows viewing the same document/notepad so I guess that functionality exists in a digital sense, even if it doesn't feel exactly the same as having your desk covered in paper!).

I've learned through much frustration that it's far better to keep input values in their own cells, and reference them in the formulas. Pretty much the only numbers hard-coded in my spreadsheets calculations now are constants in the equations out of the design spec.

Yeah really hate when people do this. All hidden away awaiting someone else to use it in error (which I've also seen to disastrous effect). Sort of related to my comment regarding people showing the end result of the calculation but none of the working, does make things hard to follow unless you have access to the spreadsheet to see what the cell actually contains. People seem to think if there is a spreadsheet for something that it must be correct, using it blindly with no verification... pet hate!

Celt83 regarding your comments on OneNote, yeah we already have the standard calculation sheet in OneNote as a background, the printing a single page thing from one OneNote page just seems to be a massive oversight for practical use. Hence my comments on the fact it's primarily a note taking app, other features are simply a bonus. Guess I'll get used to it though.

Been playing around with Word since you posted, inserting the standard calc pad page as a background image in the header, tweaking font settings you can get the spacing between lines and font size to match a 5mm grid which actually looks pretty nice, adding all the header information as bookmarks/document properties that can be populated using VBA. The inking stuff is similar to OneNote, but the fact that the text can flow onto additional pages is exactly what I was thinking I was looking for. Much more scope for automating things as well with the availability of VBA which appeals to me.

Word still seems to have the same limitation regarding inserting PDF printouts, doesn't seem to be any nice way to do it unless you take a screenshot unless I'm mistaken. I did manage to do something to import a picture of a PDF but the quality was god awful, barely able to be read on the screen, and doing the same steps again just resulted in an icon appearing despite all information on the internet saying the content was supposed to appear as an image. It's mainly getting excel printouts into the word file so you have a continuous set of calculations in one file ready to be PDF'ed that seems impossible and is my main barrier to getting things working as I expect?

Alternatively you could assemble PDF printouts/drawings into the final PDF from individual PDF's, but if you have a lot of individual PDF's it's a bit messy to manage making sure this particular printout is supposed to insert exactly at page x. All good if you're the only one working on a job, but as soon as you have others not following your workflow I see the potential for a real mess. I guess I can easily make a tool for automating the final assembly using VBA to go through a contents page of sorts in excel with all the files sitting in a directory and print each file to PDF or merge in the Excel outputs or other PDF's such as drawings. Just seems like a bit of shagging around, but fundamentally quite simple code I guess if things are organised/managed well.

The problem is the colleagues are so entrenched in OneNote it seems that it may be their way or the highway, if there was a workable solution to adding PDF's and being able to print them I probably wouldn't have even posted this topic.

One good point about OneNote is the ability of multiple people to work in the same set of calculations, none of this with Word or any other solution. Also with the move towards the non OneNote 2016 app, as I understand it you'll no longer be able to have a locally stored notebooks (Only OneNote 2016 allows this), its all hidden away on Onedrive. As I read it August 2020 was the end of life for the 2016 app, and then it was all about the version bundled with Windows 10.

Drawboard, not installed on my surface it seems (though there is something called whiteboard which may be the same), but will look into it as an alternative as well as Sketchbook even if it's just for doing some specific tasks.

Thanks for the replies so far, definitely given me some ideas.

 

[Agent666]

I believe word and one note have exactly the same equation to text feature, this one?

One note will calculate a equation if you were to write 3+3= and hit return it would enter 6 as the answer.

 

[skeletron]

As a semi-related tangent to the original topic:

For people CHECKING calculation submittals*, what is your process and what is your end goal?

*including peer review for a colleague before final submittal, or as peer review as part of a sub-contracted peer review

Are you going through the submittal line-by-line and verifying each number?
Are you identifying the final design and rationalizing it based on the assumptions?
Are you verifying the submittal versus your own design tools and methodology?


It's an important thing to consider with a submittal. Supplying reams of code can be just as confusing and irresponsible as supplying an end result page with utilization ratios. And part of your own design methodology may be your intellectual property (ie. could someone copy your calculation procedure and is this a risk to your company). I'm speaking mainly from a connection design perspective, but it could be extended to all calculation submittals in a way.

 

[Agent666]

Skeletron, I have a huge amount of experience in doing peer reviews, perhaps start another topic for more details.

Usually my approach goes something like this though, keeping in mind every project is different in terms of the calculations and documentation you received to review:-

a) Review their model, Etabs usually, get any modelling issues corrected, often there are lots of minor things, sometimes major things that completely change the answers. Sometimes this is via early involvement to prevent significant rework by the designer.

b) Sometimes an independent model is created, but this does create some issues sometimes when comparing numbers. I usually only do this if the designer is not listening to the fact that they have fundamentally modeled something incorrectly.

c) Review drawings, most design related issues will be readily apparent based on experience, I usually return a comprehensive set of comments as early as possible that require addressing written directly onto the drawings. This gives them some stuff to keep them busy while you dive into the analysis and calculations in more detail.

d) Review calculations, sometimes this is totally independent checks based on my own analysis using my own tools, sometimes it simply verifying their calculation for more simple stuff, verifying you agree with demand, verifying the spreadsheet they are using gives the same answer, etc. Review all design assumptions, usually there is a measure of asking yourself what would I do in this situation. Always reviewed against minimum code requirements, designer can obviously exceed these. If required I make suggestions regarding how they might improve details that aren't up to scratch, provide a solution you would be happy accepting for example can go a long way towards making sure you don't get a load of rubbish back a second time.

e) Usually the level of checking is dependent on your judgement and what you start to find or uncover. Usually if there are mistakes, you'll find more. If there are fundamental mistakes or errors, you'll find a whole lot more if you go looking. Keep looking until you don't find anything else.

f) I try to review almost everything, whether its my own independent checks or skimming through their calculations looking for errors in thinking or numerical errors. Judgement and experience teaches you where to look. I once passed something over eyeballing it and saying it looked ok, I had some spare time at the end of the day and decided I'd look at it anyway in detail and discovevered they'd neglected all the live load on a number of critical spans. You never can tell where the mistakes will be so it pays to be as thorough as possible.

g) Often you'll need to request more calculations for stuff that was wrong, or which doesn't seem to have even been designed (designed by drafting, or guess-timation).

h) Usually you feel like you're the ambulance at the bottom of the cliff, the last line of defense against stopping someone doing something really stupid and dangerous, usually you'd like to think even the most basic level of verification internally would have identified some of the issues you note (be they design or documentation issues). Sadly this seems to never occur to the level of rigor required. People seem to operate outside their level of experience/comfort all the time, they don't know what they don't know.

i) Sometimes you feel like you are teaching the designer how to design. Sometimes they do not appreciate this. I get hate mail sometimes, sometimes its personal, but I like to think I'm saving lives and ultimately doing my client a favour. Sometimes you get blamed for the project being delayed by everyone, but I throw it back on the designer in terms of if they had designed and documented it correctly in the first place we wouldn't be in a delay situation.

j) it's evident when people are not finished the design, these are the most painful reviews. Better to take an extra week and finish the design, rather than take 3 weeks extra to do the review. Sometimes I've spent a week documenting the information I'm missing without even checking a single number, this is not factored into my fee, but I often have zero choice but to suck it up.

k) when calculations are well laid out with supporting information it actually quite enjoyable reviewing how others approach a given situation. Though it rarely happens these days that things are that well documented. Explain what you are doing, because in 10 years time I guarantee the author of the calculations won't even be able to tell what they were doing 90% of the time. A good calculation set will almost have as many words as number explaining what is going on.

l) More often than not its a frustrating experience, you'd rather be teaching the person sitting beside you than teaching some other consultant how to be a better engineer.

m) somewhat related I recently posted this post and this post to my personal blog regarding verification and peer review, albeit related to verification of software tools with a bit of a rant on peer reviews and peoples reactions thrown in for good measure. May be of some interest regarding the topic though.

 

[Agent666]

back on topic... if I convert the PDF to an SVG file format image you can import it into word with no quality loss.

The only issue is you need to stuff around with section breaks so you can adjust margins for the page you insert it onto in word to accommodate expanding the image to the full paper size. Easy enough to automate this though via some macros I suppose (click a button an all the necessary steps are taken). But it's the only way so far I've found to bring a single page PDF into word with satisfactory results. Haven't tried multi-page PDF yet.

 

[Celt83]

Agent if your company has an active subscription with Microsoft I highly suggest getting familiar with their teams app and underlying systems. Like everyone else they, Microsoft, are pushing the cloud culture. The days of owning an application and having total control on your data appear to sadly be coming to an end, unless you are willing/able to drop out into the Linux eco system.

Open Source Structural Applications:

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

 

[Celt83]

Take a look at ghostscript for the pdf to svg conversion can do some simultaneous things there with a batch script or macro.

Open Source Structural Applications:

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

 

[dold]

Agent, I havent been able to divine the nature of your situation, but I will explain my (seemingly successful) workflow:

Who is this calculation package for (as Skeletron alluded to)? AHJ code reivewers? They can always ask you for specifics. In-house managers/etc? They can ask me, or look in the project file under ".../project/calcs/'dolds calcs'/..." Forget electronics until you've thought out your design, do that later. I take a pencil and company grid pad and spill my thought process on to the paper, in words, then numbers, diagrams, etc. State assumptions. State theories to be used. REFERENCE the standards, don't waste time copying them. If the concerned party wants to verify your code reference, they can look up section x.xx.xxx of ASCE 7-10, etc. Just write down the section as you're looking at it, dont waste your time copy/pasting it. Generally speaking this takes me about a page or two of largely spaced handwriting per "system" of the building.

Take a single story retail building as a simple example:
Roof: State your loads, state your spans and spacings. Select a joist and write down the allowable values for reference out of the SJI/vulcraft/etc book. Got some fancy snow drifts? Add another sheet with all that figured out and make sure you account for it. Got a spreadsheet for that? Print it ON THE PRINTER. Put it directly behind the physical piece of paper you just wrote on. If you get questions about the validity of said spreadsheet, be ready to validate it when the time comes. Don't waste time preempting that. Beams/girders: state the loads, draw a quick sketch, run it in Enercalc or whatever and PHYSICALLY PRINT THE REPORT FROM ENERCALC right then, immediately. Put it directly behind that piece of paper.
Do this all the way to the foundations. Loads change? Erase some numbers and redo the calc.

Old school and seemingly time consuming? Sure. Easy to follow? Yes, a hell of a lot faster for me to check than trying to suss out the cobbled together RISA model of however the hell you decided to do that, and when, and why, and since you quit I can't ask you.

Show me what your loads are. Show me the capacities. Show me the load path. Do all of this in a logical way that a human can easily understand, WITHOUT having to know how to use the software package you used.

There is nothing more time-wasting than getting a bunch of bullshit printouts from some subcontractor designing stairs/etc and having to sort through M1, M3, M35A, etc etc etc. and match it up with whatever unreadable result you have 58 pages later. Just tell me what the load is and show that the member can support it. That is what I need to know.

That said, take your stack of papers, with all of the computer parts printed out, scan it in to a compressed black and white PDF and send it.

I hammer on the "print it out" aspect because I often have to go rifling through a physical pen-and-paper project folder 6 months later and hunt down "see spreadsheet" and have no idea what the hell "spreadsheet" is.


This may be totally off base of what you were getting at, but these are some of my peeves that i've tried to avoid for the sake of redesigning an entire building later on, especially when the person cannot explain how they did it to begin with. That's a whole separate issue, though...

 

[canwesteng]

You should kill OneNote and get on Smath or Mathcad, and if you type slower than you write by hand maybe take a typing class in the evenings. Design development doesn't belong in the final calculation for checking, if you're experienced enough to do checking you're experienced enough to evaluate the design and see if it makes sense, or at least experienced enough to get off your seat and talk to the desinger. If you're using Smath or Mathcad you stand to save tons of time checking, vs Onenote or excel (shudders). You'll also find you quickly develop a library of standard calcs (concrete flexure, steel in flexure and compression, monorail design) that save you hours a job.

 

[IRstuff]

Seems to me that if you using a calculator, then you ain't doing it right anyway, since the official tool of purists is the slide rule; God forbid you're using a #@#@$#@$ programmable calculator ;-)

Time marches on; we used to spent hours getting the units conversions and scaling correct, and oft times, we'd get them wrong; modern tools like SMath or Mathcad, blows that out of the water, you can use furlongs next to meters next to microinches and NEVER get the conversions wrong. Moreover, having leftover units tells you something got screwed up, i.e., N/m vs. N-m. The other HUGE plus is not having to write or transcribe anything, i.e., no swizzling of "3.745" into "3.475," because all the calculations are live and you don't have to copy the results from one calculation to another. Not only that, but SMath, Mathcad, and a couple of others gives you EXACTLY what you would have put on paper with a pen/pencil, again with ZERO transcription errors. Below is one page from a 14-page analysis document in Mathcad that includes live calculations, units, images, and other relevant information, all fully self-contained.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Celt83]

I don't think the issue is Agent not wanting or knowing how to use these other tools, it's that the design office they are now in more or less requires the Onenote workflow.

Open Source Structural Applications:

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

 

[Celt83]

The Whiteboard app != Drawboard

The whiteboard app is just a digital whiteboard where multiple team members can actively draw, ok for working thru a detailing issue across offices.

Drawboard is a full PDF editor app with pretty decent pen support, which may actually fit for some of the workflow you describe.

As far as bringing in excel sheets/print outs or other analysis print outs your going to be hard pressed finding a good system for that, we typically just individually print everything as needed and manually assembly a compiled PDF calculation package with Bluebeam/Adobe when needed.

The one infinite page issue with Onenote is something a lot of folks do not like and have voiced the opinion on Reddit and the Onenote uservoice

per here, sounds like Onenote 2016 will get security updates thru the Office 2016 support life up to October 2020.



Open Source Structural Applications:

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

 

[fel3]

+1 for Mathcad and SMath Studio. With both programs, the math is visible and easy to follow* (unlike Excel), and units are handled beautifully (unlike Excel, which doesn't do units natively). This makes debugging and verifying a calculation procedure many times easier than with Excel. It's also easier to annotate calculations in Mathcad and SMath. Both programs provide quick and easy recalculations for new data (just like Excel). Hand calcs are visible and easy to follow*, but units require manual intervention and recalculations are time consuming.

*Assuming the engineer does a good job of organizing the calculations.​

Of the two, Mathcad looks better, but SMath is free. I have used Mathcad since the late 1980s (starting with v2.5.2, the last DOS version) and SMath for maybe ten years. I prefer Mathcad over SMath, but each program has its advantages. Also, both programs support document templates (like Excel, Word, etc.).

The beauty of electronic calculations, whatever the source (e.g. Excel, Mathcad, SMath, etc.), is that that the calculations can be reused for new projects without having to start from scratch each time. Sometimes, only the data needs changing and other times portions of the calculation procedure need to be changed, but in any case much time can be saved. The most impressive time savings come from repetitive calculations where only the data changes. On the other hand, when reusing calculations, it is imperative that the engineer verify all inputs and verify that the math procedures are correct.

Fred

==========
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[BridgeSmith]

I tried Mathcad several years ago. I found it would be easier in some aspects but more cumbersome in others, especially when dealing with large tables and the automating of repetitive calculations that I do with macros in Excel. My preference for Excel is probably mostly a matter of familiarity, though.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[IRstuff]

There are options for embedding Excel tables inside of Mathcad, which then allows you to set up the calculations relatively easily.

One significant advantage of setting up re-usable worksheets is that things don't get lost, and you won't "forget" to do some special case, since it's already in the worksheet.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[BridgeSmith]

It's a bit of extra work to write out the equations in the labels for the cells and then write the equation in the cell, referencing the values from other cells, but it works for me. I have the equation that I used shown for the checker to see what I did, without having to work with 2 programs.

Rod Smith, P.E., The artist formerly known as HotRod10