After the presentation I thought about Axels comment about losing the 3D aspect of the design. I mentioned that it was because of the constraints I made. The most important constraint is actually the contest context. This is defined by a point load in the middle of the span of the bridge. Such a load doesn’t really fit with a uniform element design. So you have to make a choice.
Constraints for a more uniform 3D design:
-Equally divided load over the whole span of the bridge.
-maximum of 4 to 6 elements attached in one node.
-lower maximum length of one element, for instance 5 centimetre.
I am more interested in the maximum structural performance of the bridge in the contest context so that is what I’m going to continue with.
Today I sharpened the model a little bit further, the upper part is more logical now.
The next days I’m going to try to find the perfect angle of the tie rods.
Constraints for a more uniform 3D design:
-Equally divided load over the whole span of the bridge.
-maximum of 4 to 6 elements attached in one node.
-lower maximum length of one element, for instance 5 centimetre.
I am more interested in the maximum structural performance of the bridge in the contest context so that is what I’m going to continue with.
Today I sharpened the model a little bit further, the upper part is more logical now.
The next days I’m going to try to find the perfect angle of the tie rods.
Thanks for posting these reflections on your presentation, Bart.
BeantwoordenVerwijderenHere is a summary of my thoughts from the notes I took during your presentation:
- the presentation was concise and - as requested - did not dwell too much on the question and hypotheses but rather emphasized the current work of modeling and testing
- the relationship between the geometric (parametric/associative) model and the analytic (FEA) model was not made clear
- the progress of the work with respect to the original aim of comparing 2D- and 3D- bridge designs (and/or of supporting the claimed superiority of 3D designs) was not demonstrated; however, it seems that this could still be done within the modeling-testing-feedback cycle which you have begun
André
Hi Bart,
BeantwoordenVerwijderenthanks for the view of the refined model, as mentioned in the presentation I think the approach may benefit from developing an alternative paralel design to the one you have that really focuses on testing the 3d nature of the structure you were interested in. That way you may be able to test that such an approach could come close to the conventional arc designs but maybe offer a lot more variety in form with little penalty in efficiency. Kristina Shea's work on assymetric doem designs was based on such a hypothesis, showing that there is a whole family of designs beside the rotational symmetric dome structures that performs very close to the optimal ones but allowing for a large range of formal expression using a rule based generative evaluative feedback loop.
http://www.google.nl/url?sa=t&source=web&ct=res&cd=10&url=http%3A%2F%2Fwww.tudelft.nl%2Flive%2Fpagina.jsp%3Fid%3D7801ecbe-cdb8-46bf-ae2a-2a14955855c9%26lang%3Dnl%26binary%3D%2Fdoc%2Fshea_delft_synth.pdf&ei=ITWxSYCyJsKP-AbH7cX2Ag&usg=AFQjCNF5hOFZ-cVJ6YgNzYZWi439Yz5Bnw&sig2=zAJlHLCxOuUO5pcGg69bdw is a link to a pdf presentation of hers she gave recently here.
So not to keep you doing from what you are developing, just a response to the question how to possibly get some of the 3d structural nature back into the project
Axel