Welcome to MCAD Central

Join our MCAD Central community forums, the largest resource for MCAD (Mechanical Computer-Aided Design) professionals, including files, forums, jobs, articles, calendar, and more.

Register Log in

Using BMX for linkage design

coseng

New member
hi,


i am trying to use BMX to design a spring linkage that has a specific leverage ratio that varies throughout the linkage's range of motion.


i've created a skeleton model of the linkage with all the measurement analysis features that define the leverage ratio of the mechanism and also made a parameter (using a polynomial expression) that defines the graph of the desired leverage ratio, all driven by the spring length dimension. the problem is i don't know what analysis to run to optimize the linkage dimensions to get the leverage ratio curve that most closely matches the desired curve throughout the desired range. is this an optimization? if so, how do i set a range for the driving dimension (spring length)? is it a multi-objective design study? how would i go about doing it?


is this possible with BMX? i also have MDO if needed.


thanks,


chris
 

Mloew

Moderator
If you have the desired motion curve, you can create a measure from the point of interest. Then take pass that measurement through an analysis feature of the mechanism simulation. Extract the maximum distance through the mechanism travel. Then create a BMX optimization feature that minimizes this value. Cool. I love minimizing the maximum! Good luck!
 

Kaz Z06

New member
Chris,<?:namespace prefix = o ns = "urn:schemas-microsoft-com:eek:ffice:eek:ffice" />


Matthew's proposal should work if I read your problem correctly. At the same time, this sounds like a classic case to try and exercise the very little known BMX function called "graph matching". Unfortunately, PTC does not include teaching this functionality any longer as part of their formal training class, but it can be very powerful. You basically need to define a "graph feature" in Pro of your desired function then create a field point that will generate a curve based on your actual measure. The field point curve is then compared to the desired curve and the dimensions you provide are adjusted t minimize the difference between the two. The trick for your problem is to turn your leverage ratio measure into a continuous curve.It's been a little while since I've used this function (and for a very different problem), but I'll give it some more thought and propose something to try. It would help me if you are able to post your model in this forum so I could take a closer look and experiement.


Regards,


Kaz Z06
 

coseng

New member
2007-03-15_132107_single-mcadpost.zip


thanks for the responses.


for matt's relpy do i need to use MDO and have an assembly? i don't have advanced modelling so am fudging skeleton models. i don't have motion paths, just graphs describing the desired leverage ratio. would this still work?


for kaz, i found a help subject on compare two graphs but can't figure out how to greate a graph of the actual leverage ratio and/or relate it to the field point.


attached is a file with the situation. R-VR-SQUARED is the actual leverage ratio and POLY-VR-SQUARED is the desired ratio curve. the driving dimension is d27, from 11.8 to 14.


thanks again for the help.


chris
 

Mloew

Moderator
Chris,

You will not need MDO, but will need MDX. MDO is only needed when forces are to be considered. If your problem is limited to kinematics and not dynamics, you can use the standard, built-in functionality of MDX. I'm downloading the model and will take a look tonight at it.
 

Mloew

Moderator
Chris,

I tried opening your file. Apparently I was expecting an assembly. Can you post your entire mechanism?
 

coseng

New member
Matthew,


The assembly is not done in mechanism so the assembly does not have joints defined. it is an old model that has gone through many revisions and has the associated baggage.


any suggestions?


thanks,


chris
 

Kaz Z06

New member
Chris,


I concur with Matthew - if you can create & posta simple skeleton assembly of the individual links with their appropriate joints we can quickly give you some advice on how to proceed. I would start with the part file you posted and copy it as many times as there are parts. Then simply delete the curves not needed for each part and assemble them together to represent your mechanism.


Cheers
 

Sponsor

Sponsor

X
Top