Re: Linear motion design Q

Chris,

Thanks for pointing out my desing flaws. %^)

I am not looking to build a machine with the higest precision
possible, but something that will perform what I'm trying to do with
it. The reason I am leaning towards CRS it that I bought some 3/4"
round stock (3Ft) and supported it at each end. I hook up a dial gauge
in the middle and preceded to press my finger in the middle and could
deflect it quite a bit. I was thinking that a piece of 1/4" x 2" CRS
place on edge (2" flat side facing verticle) with the bearing running
on the 1/4 would not deflect as much.

I will attempt a ascii artwork.

O
---------
2"
---------
O

I am trying to build as cheap as possible but with as much accuracy as
possible. I still have to purchase Acme screw with Turcite nut,
steppers, controllers. I already have a box with Debian Linux
installed (RTlinux) and EMC.

>
> What makes you think CRS will not deflect? All steel has nearly the
> same modulus of elasticity, and simply being square in cross section
> will actually hurt compared to being round, for a given mass of metal.
>
> Is the CRS perhaps a tube or channel with better geometry? That

would help.

>
> But CRS has some problems for high precision applications - it's not
> made to nearly the accuracy of shafting, being simply rolled instead
> of ground. Yes, you can buy precision ground flat and square steel
> stock, but that is neither CRS nor cheap. Also, I belive it may warp
> from internal stresses if machined (or is that HRS? I can never
> remember)
>
> You may come up with a quite nice machine based on it neverless - just
> want to point out what seems to be a flawed assumption.
>
> Chris Stratton
>
> -------------------------------------------------
> Christopher C. Stratton
> Engineer, Instrument Maker, and Horn Player
> ccs@m... 617 628 1062
> http://web.mit.edu/~stratton/www/brassbuild.html