Re: shaft torsion ?

I agree with hanermo that 1/2" round stock is way too small for this
application, both for the reason he stated and also for elimination
of the possibility of the shaft acting as a torsional spring even in
steady state operation.

Many driveline components in all types of machinery end up being
sized for stiffness rather than strength in order to achieve
satisfactory performance. The "softer" the shaft is in torsion, the
lower its torsional resonant frequency will be. If you happen to hit
the resonant frequency, then any "cogging" of the gears which may be
imperceptible under most conditions can become a driving force for
torsional vibration whose amplitude becomes greater with each cycle.
In other words, from smooth operation to a violent shudder in a
fraction of a second. By increasing the shaft diameter, you move the
first harmonic, or rpm where the shaft can resonate, upward and past
the range of normal operation.

Torsional stiffness increases roughly as the cube of diameter. So, a
1" shaft is about 8 times as stiff as a 1/2" one of the same length.
While the 1" shaft may be totally adequate for stiffness, I'd be
inclined to go with something on the order of 1.5" OD tubing with a
16ga wall for that sort of length. Torsionally, it'll be over 20
times as stiff as the 1/2" round, but it should also be large enough
in relation to its length that it wouldn't require any midspan
bearing. 5 ft of 1.5" DOM mechanical tubing would cost you the same
or less than a decent midspan pillow block, so the cost is
insignificant in the overall picture. The mandrel drawing process
creates a product that's pretty darn round and true, so it'll tend to
run smoother than something like a piece of common pipe or conduit.
You're using the OD for stiffness and whip resistance, so there's no
need for overkill on the wall thickness. Heavier walls will magnify
the imbalance effects of any runout in the tube without contributing
anything significant to the overall strength or stiffness.

I'd avoid any use of EMT as a driveshaft. It's dead soft for ease of
bending and tends to be neither round nor straight. It's good enough
to act as a wire raceway, but really not much good for anything
requiring any accuracy.

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "gcode fi (hanermo)"
<yahoog@...> wrote:

>
> Many comments on this but ...
> There is a LOT of force in accelerating and decelerating the gantry.
> The force is equal to the MAX TORQUE and POWER of your drive motors
> (steppers or servos) and it is easily hundreds of kg.
>
> The gantry must accelerate and decelerate FAST.
>
> As previously pointed out, the friction is very low and

insignificant -

> the acceleration force is definitely not.
>
> The best industrial machines accelerate at 1-2 G (which is a LOT)

and

> the biggest problem is making the gantry rigid while keeping weight

down.

>
> Typical homebrew machines use about 3 Nm motors - at 1:1 on timing

belts

> they push at about 100 kg laterally, ALL of which goes into and IS
> NEEDED to accelerate and decelerate the gantry properly. And its

still

> one half to one third of industrial machines.
> There are a lot of machines done over at mechmate.com - all of them
> support these numbers.
> There are a lot of industrial machines and data - all of them

support

> these numbers.
>
> You need the acceleration for clean corners and small details,

where the

> gantry must accelerate/decelerate as fast as possible.
>
> 0.5" is way small.
>
> Make you choices from what works for others ?
> Good luck,
> h-
>