craig1410
|
| posted on 6/9/03 at 05:34 PM |
|
|
Solidworks tips needed - Please help!
Hi,
I know this probably isn't the right forum for this question but I the people who can probably answer my question will most likely appear here
(hopefully anyway...)
Okay, Solidworks...
I have been messing about with Solidworks trying to design my chassis and although I have gone through all the tutorials which come with it, and read
the help text extensively, I feel I am missing the point somewhere when it comes to mating (ooh err missus). For those of you who don't use
Solidworks mating is all about getting the various parts which you've designed to stick together correctly. Much more difficult than the other
kind of mating, believe me... 
Anyway, how do you go about getting everything to go together without getting mating errors because they are driving me crazy!!
I get the feeling that I should have some sort of master sketch which locates all the parts and the mates will be between the edges of the parts and
the master sketch rather than between the parts themselves. I have got as far as modelling the base rails for the chassis but as soon as I start to
place the "H" tubes on it all goes wrong.
Can someone out there please let me in on the secret design process or send me a file with a chassis or part of a chassis already built. I have been
designing my individual tubes as per the McSorely plans and I have all the angles cut in the ends etc but just can't seem to stick more than a
few of them together before I get mating conflicts.
I'd like my CAD drawing to catch up with the physical design so that I can start to get the benefit of it during the next stages of the
build.
Any help will be much appreciated!!
Cheers,
Craig.
|
|
|
|
|
Alan B
|
| posted on 6/9/03 at 06:49 PM |
|
|
Basically, solidworks is shite....
Use mechanical desktop instead..
There you go that should get you some response...
I'd suspect the mating errors are due to conflicts caused by minor inacurracies or rounding.....e.g. you may have a pair of parallel tubes and
you are fititng one between that is angled each end...it may be that there is a discrepancy of even less one-ten thousandth of an inch, but it is
enough to show a conflict......
Perhaps there is a setting you can change the mating tolerances?....
Anyway, someone knowledgeable is bound to come along...
|
|
|
craig1410
|
| posted on 6/9/03 at 07:11 PM |
|
|
Thanks Alan, I appreciate your opinion although perhaps some may not...
I agree that the mating errors will be due to rounding or something as when angles are involved I will undoubtedly get wierd dimensions. I'll
look for some sort of allowable discrepancy setting as you suggest.
Thanks,
Craig.
ps. Alan's point may be valid that Solidworks isn't the correct/best tool for what I'm trying to do. Does anyone else have a
suggestion on what tool is best for Locost designing? Ideally a shareware or non-commercial license or other low cost product.
|
|
|
Alan B
|
| posted on 6/9/03 at 09:28 PM |
|
|
Actually Craig I'm certain Solidworks is fine. (My mate Jim does OK...)
As is Mech desktop, Inventor, Solidedge, Pro-E and many others.
Personally, I often don't use mating features...I'll often just say line up the corner of a tube with another, or however it joins.
I find assy mating more useful in parts that move rather than in plain structures.....again just IMO..
|
|
|
Alan B
|
| posted on 6/9/03 at 09:34 PM |
|
|
Example for you Craig......
No assembly mates in this......
Just stuff positioned where I wanted it.
|
|
|
Viper
|
| posted on 6/9/03 at 09:48 PM |
|
|
so whats wrong with a pencil and paper??
Tut,,i dunno youth of today
|
|
|
Alan B
|
| posted on 6/9/03 at 10:18 PM |
|
|
quote:
Originally posted by Viper
.......i dunno youth of today
Hey cheers mate..
I better make that two beers for you..
|
|
|
craig1410
|
| posted on 6/9/03 at 10:19 PM |
|
|
Viper,
I didn't realise you were an old fart!
I did O-Level Technical Drawing at school actually and used to love drawing isometrics. I still quite often doodle isometrics when sitting down
thinking about how to build stuff and find it useful to get ideas together in my mind. However, to figure out how things really go together accurately
I think CAD is really useful.
Alan,
Thanks for the tips, I think that you are absolutely right that things should just be placed where you want them to go so I guess a master layout
sketch is what I need to create and then just align the bits with that sketch. I would like to use mating surfaces to allow me to move my suspension
around though to check for clearance etc.
Anyway, thanks guys!
Craig.
|
|
|
Viper
|
| posted on 6/9/03 at 10:44 PM |
|
|
I am not an old fart
i just act like one
|
|
|
blueshift
|
| posted on 7/9/03 at 04:17 PM |
|
|
Alan, do you mean AutoCAD Mechanical Desktop? *shudder*
I tried using that before Solidworks and it was horrific. Solidworks is a walk in the park by comparison.
Mates are the way to go imo, I plan to redo our chassis design as an assembly where parts are defined by the assembly and their meetings with other
parts and assembly dimensions.
The point being I can then do nice things like add +4 to the width of the chassis and all the members will recalculate their lengths and angles and
still be connected up right. Then I can just print out the cut list and off we go.
Mates are also fun for testing out the suspension.. I did a working model of a coilover shock, it had about 50 mates in it and took about 5 seconds
for each frame of movement of the assembly on my measley computer.. worked though
It has collision detection and so forth too, if we can (be bothered to) do sufficiently good models of the engine, diff, other parts etc we can see if
they're going to fit together further down the line before we weld things in (like the transmission tunnel)
Should be able to model in any funny brackets we need and email them off to my friend with a laser
On the other hand I might get bored of all that monkeying around and just go and bash the metal by trial and error.
A nice fully featured cad model of a locost with mechanicals might be useful to others though, I know Jim McSorley's ones have been, so
I'll see what we can do..
|
|
|
Alan B
|
| posted on 7/9/03 at 04:59 PM |
|
|
I think it is a case of once you get into one particular solid modeller, then others seem "odd" or lacking....what you get used to I
guess....
I have used the mates a lot in inventor and found them useful.....but so far still prefer mech desktop overall....maybe I used AutoCAD too
long...
All have their good points but none are perfect......
I solid model all day, every day (except when I'm on here..) so I have to use something I'm comfortable with...
|
|
|
craig1410
|
| posted on 7/9/03 at 05:16 PM |
|
|
Blueshift,
What's the secret of getting more than a few mates to work without getting mating conflicts? Any chance of giving me some hints or even better
sending me a file with an example of what to do?
I built my lower chassis rails by defining each tube individually and then mating them together and it worked up to a point but eventually it just
fell down due to mating conflicts. Is there a rule of thumb which says which types of mate to use? I was using a combination of face, perpendicular,
parallel, vertex, point and edge mates but didn't really have an overall strategy which is what I think the problem might have been.
Any advice you could offer would be very much appreciated!
Cheers,
Craig.
|
|
|
blueshift
|
| posted on 7/9/03 at 11:56 PM |
|
|
Taken this to email so as not to waste anyone else's time.. if anyone other than craig is interested, let me know.
|
|
|
kingr
|
| posted on 8/9/03 at 11:14 AM |
|
|
Just to make use of the obvious cheap gag :
I can't get any of my mates to work, because they're all lazy gits.
There, had to be done, if I hadn't Steve G would have beaten me to it.
Regards,
Kingr
[Edited on 8/9/03 by kingr]
|
|
|
mackie
|
| posted on 9/9/03 at 10:19 AM |
|
|
Yeah, as you can see from my avatar I've been playing a bit, based on the McSorley plans. Basically some bits wont mate, but you can still have
a "fully defined" model where non of the bits can move, it's just that they'll only be fixed at 1 end. As blueshift said
he's wanting to make it work a bit more dynamically that that but i mostly treated the chassis model as a learning experience anyway.
|
|
|
craig1410
|
| posted on 9/9/03 at 12:49 PM |
|
|
Thanks guys, Blueshift and Mackie have given me some good advice which I'll try to put into practise as soon as I get some free time. It seems
that the order of the day is to try to avoid any "mating loops" (ooerr again!) and just try to have everything connected at one end
instead.
If I get the chassis designed okay then I'll make the plans available on my website for anyone else to use as a starter. I hope to start
building up a library of parts like wishbones, sierra diff, gearbox etc etc in time which I will also provide.
Cheers,
Craig.
|
|
|
mackie
|
| posted on 9/9/03 at 01:35 PM |
|
|
I will watch this space since we are basically doing the same as you
A parts library would be excellent, i think i'd like to have a go at modelling some parts too. Maybe we could combine our efforts?
|
|
|
craig1410
|
| posted on 9/9/03 at 03:38 PM |
|
|
I'm all for that although to be fair to you guys I am not as competant so it may be a bit one sided to start with. I'm a fast learner
though...
As you said, watch this space!
Cheers,
Craig.
|
|
|
