Mattheck triangle method for near optimal spline without FEM

Have some feature requests, feedback, cool stuff to share, or want to know where FreeCAD is going? This is the place.
Forum rules
Be nice to others! Read the FreeCAD code of conduct!
Post Reply
Stefano4000
Posts: 7
Joined: Thu Feb 20, 2020 10:01 pm

Mattheck triangle method for near optimal spline without FEM

Post by Stefano4000 »

With this triangle method a fast optimization of loaded arcs is possible.
In Notches the normal circular arc is not the optimum.

With triangle method some points were calculated for a very good spline curve for optimal Stresses by loads.
See
http://bibliothek.fzk.de/mattheck/matth ... mmlung.pdf

See links in this postersammlung in Englisch or german with triangle method against high stresses in notches.

https://publikationen.bibliothek.kit.ed ... 12/3814714
https://publikationen.bibliothek.kit.ed ... 88/3815285
https://publikationen.bibliothek.kit.ed ... 86/3815283
https://publikationen.bibliothek.kit.ed ... 96/3814952
https://publikationen.bibliothek.kit.ed ... 89/3815520
thschrader
Veteran
Posts: 3129
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Mattheck triangle method for near optimal spline without FEM

Post by thschrader »

Interesting construction method.
Thanks for sharing.
tree_notch.FCStd
(15.84 KiB) Downloaded 25 times
notch.JPG
notch.JPG (22.29 KiB) Viewed 1047 times
thschrader
Veteran
Posts: 3129
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Mattheck triangle method for near optimal spline without FEM

Post by thschrader »

My daily bread:
2 Flanges under momentum-loading M=52 kNm. Material steel S235.
System A, welded connection: sigma > 235 MPa
System B, bionic: sigma < 235 MPa! and the flange-plate is much stiffer. Nice.
The question is: how to produce part B?
A little bit more material is not the problem, but how to weld?

@Stefano4000: I had a look at your website and the pdf, interesting stuff.
flanges_vMises_Stress.JPG
flanges_vMises_Stress.JPG (85.39 KiB) Viewed 1019 times
Stefano4000
Posts: 7
Joined: Thu Feb 20, 2020 10:01 pm

Re: Mattheck triangle method for near optimal spline without FEM

Post by Stefano4000 »

Mit einer Drehbank kann man den Spline nachfahren und damit die dicker ausgeführte Schweißnaht optimal nacharbeiten.

Man schneidet jedoch das Gefüge der Naht an und damit nicht optimal aus anderer Sicht sein kann.

Bei dickeren Bauteilen hat man die Naht nur als ein Teil des Splines und damit des Kerbverlaufs.

Im Beispiel als Flansch ist der Spline durch die konstruierten Punkte nicht realisiert. Der Spannungsverlauf hat als polygonale Ecke noch Potential.

Eine Funktion in Freecad zur Erstellung dieses Splines mit Anfangs-, Eckpunkt und Endpunkt hatte ich mir gedacht.
Eine zweite Funktion für die „Stufe“ mit 2 Eckpunkten und 2 Splines wäre auch ganz nett.
UR_
Veteran
Posts: 1354
Joined: Tue Jan 03, 2017 8:42 pm

Re: Mattheck triangle method for near optimal spline without FEM

Post by UR_ »

thschrader wrote: Fri Dec 11, 2020 4:06 pm The question is: how to produce part B?

https://www.vhg-gruppe.de/vorschweissfl ... n-2633pn16

Screenshot 006.png
Screenshot 006.png (87.03 KiB) Viewed 987 times

Probably not far away from best fit :roll:
thschrader
Veteran
Posts: 3129
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Mattheck triangle method for near optimal spline without FEM

Post by thschrader »

UR_ wrote: Fri Dec 11, 2020 5:49 pm ...
Probably not far away from best fit :roll:
Sorry, far away from best fit ;)
This connection does not work for high momentum loading
(the gap between the flanges at the bore hole).
We had diskussions with our customers/controlling engineers long ago about this kind of design.
No chance.
thschrader
Veteran
Posts: 3129
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: Mattheck triangle method for near optimal spline without FEM

Post by thschrader »

Stefano4000 wrote: Fri Dec 11, 2020 5:38 pm Im Beispiel als Flansch ist der Spline durch die konstruierten Punkte nicht realisiert. Der Spannungsverlauf hat als polygonale Ecke noch Potential.
Eine Funktion in Freecad zur Erstellung dieses Splines mit Anfangs-, Eckpunkt und Endpunkt hatte ich mir gedacht.
Eine zweite Funktion für die „Stufe“ mit 2 Eckpunkten und 2 Splines wäre auch ganz nett.
Using splines in sketcher is tricky. You cannot define define tangential connections (at the moment).
Thats why I used the polygon as a first step.
I have seen a picture of a fish with your "Zugdreieck" (tension-triangle). Can you explain how to
use this with CFD? FreeCAD has a CFD-workbench, do you know that?
Stefano4000
Posts: 7
Joined: Thu Feb 20, 2020 10:01 pm

Re: Mattheck triangle method for near optimal spline without FEM

Post by Stefano4000 »

This triangle method is good for all gradient problems at edges and notches in fem and Cfd.

In Cfd the problem of optimum is more complex from laminar to turbulent. This triangle construction is very good with low reynolds numbers of fluid.
For full turbulent cases with great complexity it is a good start point for parametric optimization of contour points of this spline.
Stefano4000
Posts: 7
Joined: Thu Feb 20, 2020 10:01 pm

Re: Mattheck triangle method for near optimal spline without FEM

Post by Stefano4000 »

In the example tree_notch you can make a 2D-Cfd calculation and the spline contour does have a significant better pressure drop value than a sharp or circular design.

So inflow is up and 2 outflows are at the bottom left and right.
Laurie Hartley
Posts: 522
Joined: Mon Feb 27, 2017 5:33 am
Location: Australia

Re: Mattheck triangle method for near optimal spline without FEM

Post by Laurie Hartley »

thschrader wrote: Fri Dec 11, 2020 4:06 pm The question is: how to produce part B?
A little bit more material is not the problem, but how to weld?
Did you figure out how to weld it??? :)
Post Reply