First look at a combined point load and distributed load. I’m not exactly happy with the shear diagram where it makes the vertical jumps. I will need in to add in a correction factor (additional point) at each point load and internal support so that the shear is properly reported at those locations.
1 Like
Still putting out a few fires and bugs. I had a small bug in my conditionals where loads were symmetric (like the example below), but I’ve resolved that now. My next fix is the vertical jump issue(s) with the shear diagram. I have some ideas on how to address this, just need to test it and see if it is the solution. Here is a symmetric point load scenario and the deflections for the two load cases:
P1 = 500 lbs (D)
P2 = P3 = 100 lbs (D) 500 lbs (L)
2 Likes
Overhanging beams with point loads now check out. Once again ChatGPT to the rescue to help debug my syntax and even debug my actual algorithms. This AI stuff is getting crazy good, sometimes it makes mistakes but then it is able to reason and catch itself, it’s uncanny.
Now I just need to debug for uniform distributed loads on overhanging beams. Then it is on to the actual engineering portion (AWC stuff for wood) and some final formatting of the PDF/HTML report.
I’m also not entirely satisfied with the clunkiness of the tools used to move and create the supports, some improvements on this end are needed. A load/support copy tool would be really nice, rather than having to create completely new loads and supports from scratch.
3 Likes
First look at partial UDLs with overhangs:
So with that I think the basic nuts and bolts of the matrix analysis engine is in place and functioning pretty much as expected. Of course it will probably be a few more days or even weeks before I am able to put out every little fire that may be burning undetected thus far, but we will see.
Now I will turn my attention to the following items on the todo list:
1.) Try to fix the truncation in the shear graph so that vertical jumps actually are vertical.
2.) Add in the standard engineering checks for wood beams (glulam, lumber, timber, LVL, SCL and PSL)
3.) Finish the formatting and layout of the HTML report. I may also include an option between a condensed report and a detailed report (or that may come later).
Things that are not specifically on the todo list but are interesting:
- Add in fixed and partially fixed supports, currently every support is assumed pinned.
- Engineering for steel beams
- Trapezoidal distributed loads
- Moment loads
P.S.
And to think all of this runs inside of SketchUp. 
SketchUp isn’t just a drawing program my friends, it is anything you want it to be…
3 Likes
Version 0.8.1 - 09.07.2025
- Developed the matrix analysis engine for the beam calculator using the stiffness method.
- Added a load diagram to the beam report.
- Added shear and moment graph to the beam report.
- Added a deflection graph to the beam report.
The engineering report is still not complete however by rolling this beta release I can allow potential users of this plugin the ability to test it out and assist in the debugging. The plugin can be directly downloaded from this link:
http://design.medeek.com/calculator/sketchup/medeek_engineering_ext.rbz
2 Likes
In the global settings I will be adding in the option to toggle on or off the text showing the loads and reactions in the load diagram:
As you can see it does have the potential to get a little busy but I think it would be useful to have this available as an option.
There will also be a table below the diagram showing the details for each load and each support, so the information is a bit redundant in my opinion. Thoughts?
1 Like
Version 0.8.2 - 09.09.2025
- Enabled loads and reactions (values) within the load diagram.
- Fixed a bug with end supports that are not centered on the start or end of the beam.
- Improved the formatting logic (SVG and HTML) for the load diagram.
Working on assembling the actual engineering part of the report now, adjustment factors, shear, moments, deflection and bearing checks.
1 Like
First look at the bearing check:
I also need to update my disclaimer, the disclaimer shown is for my previous online calculator, it needs some changes.
2 Likes
Spent the last two days adding in some additional code so that the vertical jumps in the shear graph (at point loads and supports) are actually vertical. It was a bit more complicated than I originally bargained on but I think I finally have it figured out:
The code seems fairly robust but tomorrow I will throw the kitchen sink at it to see if I can find any weaknesses in the algorithm.
I have’t been posting much lately but that is because I’ve had my head buried in the code. Most of this engineering code is completely new (not my typical plugin stuff) so there is no refactoring old code or any other shortcuts I can take. Some of the old beam calculator is relevant however since it was so limited in its application I’m kind of on my own with this new calculator.
2 Likes
So I’m a little stuck on the deflection portion of my code right now. For each load case/combination (13 possible combos for ASD) that is present I calculate the full Moment graph which I am then able to generate the deflection graph for each load combo. I can then easily check each array of numbers and determine which load case has the highest deflection.
However this is the TL deflection (Total Load).
To determine the LL deflection do I just algebraically subtract the dead load deflection (DL) from the max. TL deflection load case and then take the max. value?
Or do I need to compute the (array) LL deflection for each load combo/case and then look at each one to determine the max. LL deflection value?
I think I know what to do but these sorts of questions are what keep me up at night.
1 Like
I’m also adding in a deflection graph just for the Live Load, so now there are two deflection graphs/diagrams:
I always figured the vertical step is a bit of an abstraction if you think about it. If you have, say, an intermediate support wall that’s 3 1/2" of bearing, the total reaction could be seen as a UDL distributed over that 3 1/2", so the graph wouldn’t be a true vertical line, but a really steep ramp over a really short distance.
1 Like
In reality you are 100% correct. If these point loads and reactions really were infinitely sharp point loads or reactions they would cut through the beam like a knife through hot butter. A more technically accurate way to model them is as you suggest a UDL over the bearing surface and for point loads one should actually specify the bearing length for each point load, so it also becomes a UDL as well.
There really are no such thing as points loads, but that is how we like to model and think about them. I wasn’t originally too worried about them not being completely vertical but other engineers critiqued my graphs and said they looked sloppy if they were not shown as the theoretically correct vertical jump so I acquiesced.
I’m still working on the information aspects of the report as well as the bending and deflections but here is a quick snapshot of what I have so far for an example beam. Disregard the pink shaded text on page 5 (filler text for the upcoming deflection section):
Here is the actual beam drawn in SketchUp (Medeek Wall) with the loads and supports shown:
1 Like
Here is the L/d given for both the LL and TL diagrams:
However, in this case I have an overhanging beam (right side only) and I recalled that with overhangs the L/d was usually calculated differently (from previous engineering projects working with decks). Sure enough when I ran a quick test in Forte I noticed the 2L/d overhang criteria being used, see sample below:
Where in the code (ie. IBC, ASCE etc…) is this actually codified, to use twice the actual span for overhangs? I can’t seem to find it but it does appear to be common practice.
On a slightly different note you will notice that I am giving quite a bit more information than the standard report that will be produced with other comparable structural software (ie. Forte, BC Calc, EnerCalc etc…) However what I am planning to do is add another option which allows one to switch between the detailed and brief report. The idea being that the brief report will fit on one or two pages while the detailed report will probably span six to seven pages. Once I have everything coded for the detailed report it is just a matter of pairing it back to created the brief report.
1 Like
The deflection section is fairly basic but it does specify the span used to calculate the L/d as well as the x location and load combination:
2 Likes
I think it would be kind of interesting if the user could specify which analysis they wanted (ie. Euler-Bernoulli vs. Timoshenko).
To that end I need to work out the FEMs for a Timoshenko beam and also the stiffness matrix k. ChatGPT gives me this:
I can’t seem to find much information on this sort of thing. Most structural analysis texts only deal with the Euler beam since that is much more standard.
I have a book, Elements of Strengths of Materials, 5th Edition by Timoshenko & Young. I never did much with it. Chaper 5 & 6 are on stresses in beams, and chapter 8 is on statically indeterminate beams.
1 Like