Several pictures in the gallery section are described as parametric designs. Can somebody please tell me what this is means!! And does it have to do with algorythms?
i don't think this is what you are looking for but i came across this site with a thing called parametric design engine, encouraging the people to design their own homes. just click on picture to take you through the sales pitch. weird but parametric nevertheless :) link
More complex forms of CAD are solid modeling and parametric modeling, which allows objects to be created with real-world characteristics. For example, in solid modeling, objects can be sectioned (sliced down the middle) to reveal their internal structure. In parametric modeling, objects have meaningful relationships with each other (a door must be on a wall, not the floor; holes cannot be drilled too close to the edge, etc.).
Well here is his site and yes it has to do with algorithms ang scripts in architecture. If you do a research you'll find this argumet discussed many times here.
Feel free to any questions.
parametric refers to parameters, i.e, user adjustable quantities. in parametric models the designer chooses objects and then adjusts parameters. for example, a door can be placed in a wall and then a parameter such as the height can be changed from 6'8" to 7'2"
that's how i understand it, but admittedly, i'm not completely sure what is being implied by the "parametric architecture" that you are referrencing. in some ways, architecture has always been parametric in that we constantly adjust the particulars of any design in order to fit client whims, building codes, etc. it's only recently that computers hae begun to mimic this design behavior.
parametric in this context is about the generation of forms (or if you like we can call it architecture) that are generated by other means of normal modelling. For ex. a cell module is repeated progressively thru a given path by a script or algo and being adapted to envrionment or a other inputs, Thus changing a parameter or iteration of detail of mesh generates a near-completely different result. Maybe it's a blunt explanation but in terms of understanding it's like this it works. Of course there's lotsa theory about it and you'll be able to read til vomit, til time you'll experiment by yourself!
I am studying Product Architecture at Stevens and we do quite a bit of parametric modeling. Yes, it does have a lot to do with algorithms, but the output hardly needs to resemble the Columbia Studios.
We have yet to get into PM thru writing our own scripts(althought the second year students do tons of it), but rather thru using parametric modeling applications (CATIA and SolidWorks). In these programs just about anything can be established parametrically. Dimensions of a sketch. Height of that skech when extruded. Parameters are hardly limited to dimensions one could say 'extrude to the bottom face of that surface'. In that case if the location of the above surface changes then the length of all attached extrusions update with it.
Each of these examples is quite simple and of course the tip of the iceberg. Parameters can be written to have an out come of integers (risers per vertical length of a stair) angle, distance, area, volume, density, modulus of elasticity, really any dimensional, or physical property.......
Of course you can get into writing scripts with VBA in CATIA here you would be writing a protocol that would automate a series of steps over and over. The rationalization of complex forms and processes is unlimited.
Remember the primary difference between what we are doing at Stevens versus what I have read about the Columbia Studio is that we are always concerned with the manufacturability and performative issues of our designs. Actual architecture not just form finding. What a novel idea.
I am getting the feeling that some schools and schools of thought are going to start to treat the execution of a script as the new pause button in animation software. Type something in let it run see what pops out. Do that until it looks beautiful enough then start to reverse engineer it.
In CATIA the program won't even let you make a form that can't possibly be manufactured. The disussions at Stevens between creating and making is never seperated for this reason.
Use parametrics responsibly. Many other industries have been for some time. Please don't let the potential of parametrics fall into the hands of the evil algorithmic form finders. That would be a huge waste.
Thanks for all the responses and links!
Still not sure I understand everything entirely. Are there any books or even essays any of you could recommend?
Kristix, the mh link and Columbia stuff looks great. Yet why does the "parametric" part mean that they all look so blob-like? Popeye, what does your work look like? Can you post any pictures?
No, parametric design doesn't automatically give blobs. Parametric design is simply to tie numerical representations and logical relationships to shapes, just like in popeye's extrude example. Or, for example: the parameters R, M, and N decide three different radii along the axis of a vase - something which you have in each 3d modelling program. For added complexity, you could then add other parameters like a condition which says that R+M+N is smaller than a specific number, or that M = R+N / 2 etc. etc.
Algorithmic design, on the other hand, would take the three values R,M and N and feed them into a more complex iterative (or recursive) calculation, like calculating if the vase breaks if you would drop it. The result is then evaluated and the parameters R, M and N, are then fed back into the algorithm, but slightly changed according to some clever rule. All this is in order to find the "optimum" shape. One technique is "genetic algorithms" which mimics the process of natural selection - interesting but very difficult to control! I think this is why there are only a handful of interesting projects, most of them dealing with optimum structural solutions.
Hello a-f,
Thanks for the detailed explanations. Would be very interested in seeing some images made by parametric processes that are not blob-like (like the structural solutions you mentioned) Which "interesting projects" where you thinking of?
Thanks in advance!
Mark Burry is a architect and professor whose big thing is parametric design. He has been in charge of Gaudi's Sagrada Familia project for a number of years.
? Parametric Architecture
What is "parametric" architecture???
Several pictures in the gallery section are described as parametric designs. Can somebody please tell me what this is means!! And does it have to do with algorythms?
Thanks!
i don't think this is what you are looking for but i came across this site with a thing called parametric design engine, encouraging the people to design their own homes. just click on picture to take you through the sales pitch. weird but parametric nevertheless :)
link
this may help
Answers. com - parametric
More complex forms of CAD are solid modeling and parametric modeling, which allows objects to be created with real-world characteristics. For example, in solid modeling, objects can be sectioned (sliced down the middle) to reveal their internal structure. In parametric modeling, objects have meaningful relationships with each other (a door must be on a wall, not the floor; holes cannot be drilled too close to the edge, etc.).
Well here is his site and yes it has to do with algorithms ang scripts in architecture. If you do a research you'll find this argumet discussed many times here.
Feel free to any questions.
parametric refers to parameters, i.e, user adjustable quantities. in parametric models the designer chooses objects and then adjusts parameters. for example, a door can be placed in a wall and then a parameter such as the height can be changed from 6'8" to 7'2"
that's how i understand it, but admittedly, i'm not completely sure what is being implied by the "parametric architecture" that you are referrencing. in some ways, architecture has always been parametric in that we constantly adjust the particulars of any design in order to fit client whims, building codes, etc. it's only recently that computers hae begun to mimic this design behavior.
parametric in this context is about the generation of forms (or if you like we can call it architecture) that are generated by other means of normal modelling. For ex. a cell module is repeated progressively thru a given path by a script or algo and being adapted to envrionment or a other inputs, Thus changing a parameter or iteration of detail of mesh generates a near-completely different result. Maybe it's a blunt explanation but in terms of understanding it's like this it works. Of course there's lotsa theory about it and you'll be able to read til vomit, til time you'll experiment by yourself!
I am studying Product Architecture at Stevens and we do quite a bit of parametric modeling. Yes, it does have a lot to do with algorithms, but the output hardly needs to resemble the Columbia Studios.
We have yet to get into PM thru writing our own scripts(althought the second year students do tons of it), but rather thru using parametric modeling applications (CATIA and SolidWorks). In these programs just about anything can be established parametrically. Dimensions of a sketch. Height of that skech when extruded. Parameters are hardly limited to dimensions one could say 'extrude to the bottom face of that surface'. In that case if the location of the above surface changes then the length of all attached extrusions update with it.
Each of these examples is quite simple and of course the tip of the iceberg. Parameters can be written to have an out come of integers (risers per vertical length of a stair) angle, distance, area, volume, density, modulus of elasticity, really any dimensional, or physical property.......
Of course you can get into writing scripts with VBA in CATIA here you would be writing a protocol that would automate a series of steps over and over. The rationalization of complex forms and processes is unlimited.
Remember the primary difference between what we are doing at Stevens versus what I have read about the Columbia Studio is that we are always concerned with the manufacturability and performative issues of our designs. Actual architecture not just form finding. What a novel idea.
I am getting the feeling that some schools and schools of thought are going to start to treat the execution of a script as the new pause button in animation software. Type something in let it run see what pops out. Do that until it looks beautiful enough then start to reverse engineer it.
In CATIA the program won't even let you make a form that can't possibly be manufactured. The disussions at Stevens between creating and making is never seperated for this reason.
Use parametrics responsibly. Many other industries have been for some time. Please don't let the potential of parametrics fall into the hands of the evil algorithmic form finders. That would be a huge waste.
I'm affraid it already is in hands of evil formlusters! Maybe me too is one of them, but these one are goregous! Beautiful like you said!
http://www.arch.columbia.edu/gsap/54726
http://www.arch.columbia.edu/gsap/54727
http://www.arch.columbia.edu/gsap/54728
It's interesting what you do there at Stevens. Can you put a link or site of things done by students?
We are still putting the website together. I would look at the student blog.
<----
Thanks for all the responses and links!
Still not sure I understand everything entirely. Are there any books or even essays any of you could recommend?
Kristix, the mh link and Columbia stuff looks great. Yet why does the "parametric" part mean that they all look so blob-like? Popeye, what does your work look like? Can you post any pictures?
No, parametric design doesn't automatically give blobs. Parametric design is simply to tie numerical representations and logical relationships to shapes, just like in popeye's extrude example. Or, for example: the parameters R, M, and N decide three different radii along the axis of a vase - something which you have in each 3d modelling program. For added complexity, you could then add other parameters like a condition which says that R+M+N is smaller than a specific number, or that M = R+N / 2 etc. etc.
Algorithmic design, on the other hand, would take the three values R,M and N and feed them into a more complex iterative (or recursive) calculation, like calculating if the vase breaks if you would drop it. The result is then evaluated and the parameters R, M and N, are then fed back into the algorithm, but slightly changed according to some clever rule. All this is in order to find the "optimum" shape. One technique is "genetic algorithms" which mimics the process of natural selection - interesting but very difficult to control! I think this is why there are only a handful of interesting projects, most of them dealing with optimum structural solutions.
Hello a-f,
Thanks for the detailed explanations. Would be very interested in seeing some images made by parametric processes that are not blob-like (like the structural solutions you mentioned) Which "interesting projects" where you thinking of?
Thanks in advance!
Mark Burry is a architect and professor whose big thing is parametric design. He has been in charge of Gaudi's Sagrada Familia project for a number of years.
http://www.sial.rmit.edu.au/People/Professor_Mark_Burry.php
A really interesting site and projects
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.