The one aspect of this project that still bothers me the the needed accuracy of 0.1 mm/6000 mm = 0.004"/236" (because a lot of lead screws are described in 0.xxx"/inch) = 0.000017"/inch. A common value of "precision rolled" leadscrews is 0.0001"/inch, which is an order of magnitude worse.

I wonder whether one might shift priorities and just just a stepper (or similar) motor or wire to drive the cart while using a feedback loop and accurate distance measurement device for positioning?

For about $400 or less, one can get a Fluke laser ranger finder with 1-mm accuracy up to 100 m (http://www.myflukestore.com/p12884/fe_424d.php). Here is a link (one of many) that discusses the three ways to measure distance with a laser: https://www.researchgate.net/public...Finding_Methods_by_Simulation_and_Measurement

Phase shift is considered "inexpensive," but I could not find easily what that meant in dollars. The Fluke doesn't describe whether it is TOF or phase shift. The Fluke description does mention using two lasers, so it may be based on phase shift.

Of course 1-mm doesn't meed the specification you may need, but getting to 0.1 mm might be outside a reasonable budget for a target that is constantly moving back and forth.

John

 

Ahh, now I see it. You would still have the same issue as the initial setup as you would need to make sure that the wire winds in one layer or the diameter would change.

Exactly, that's the point of this entire project.

 

The one aspect of this project that still bothers me the the needed accuracy of 0.1 mm/6000 mm = 0.004"/236" (because a lot of lead screws are described in 0.xxx"/inch) = 0.000017"/inch. A common value of "precision rolled" leadscrews is 0.0001"/inch, which is an order of magnitude worse.

I'm sure it can be done, at least for an accuracy of less than 0.25mm in 6.5m. The reason I'm aiming for 0.1mm is because I've already done it in a different mechanism, and the end result was about ±0.2mm in a 7m length. Besides, if I aim for ±0.2mm, I know that in the end I'll get double that, due to elasticity and inertia in the motion mechanism that might over or undershoot one step count when motion comes to a stop.

 

I am not disagreeing with your approach, but I am just a little skeptical of the accuracy you will achieve without some other controls. For example, the coefficient of linear expansion for 304 stainless is listed as 9.6 microinch/inch•F°. A 6-m length (236") will change about 0.002"/°F.

John

 

I am not disagreeing with your approach, but I am just a little skeptical of the accuracy you will achieve without some other controls. For example, the coefficient of linear expansion for 304 stainless is listed as 9.6 microinch/inch•F°. A 6-m length (236") will change about 0.002"/°F.

John

That might be an issue, yes. But installing a digital thermometer in the thing would be quite easy for me to do. And then I'd take the corrective measures.
On the other hand, this machine is for die-punching rolled steel profiles, whose length would also be affected by the current temperature. In the end, the temp error will amount to the difference in expansion between stainless steel and carbon steel.

 

Then music wire, which you mention as an alternative, might be preferred as I recall its coefficient of expansion is closer that of rolled steel. Music wire is already cold worked, which might also be an advantage over 304.

John

 

Then music wire, which you mention as an alternative, might be preferred as I recall its coefficient of expansion is closer that of rolled steel. Music wire is already cold worked, which might also be an advantage over 304.

John

Actually, I did mention that I'd be using 316L SS. Not that it's of any consequence, thermally speaking, since its temp coefficient is almost the same as that of 304. But it is a little stronger.

But yes, I'll try to find music wire instead.

 

Then music wire, which you mention as an alternative, might be preferred as I recall its coefficient of expansion is closer that of rolled steel. Music wire is already cold worked, which might also be an advantage over 304.

John

The stainless wire of either alloy is also cold worked, all three are cold drawn to size in jewel(sapphire?) draw buttons on a draw bench.

 

SS304:


Music wire:


73X10^3 vs. 250X10^3 is no contest. That is what I meant.

John

 

I have not seen the device that this will be connected too, but do need to use a wire method? Could you just not use a rubber coated wheel connected directly to the encoder sitting on the moving part that runs along a rail or something simular?
Think trundle wheel. That would remove the stretching of the wire part.

 

Could you just not use a rubber coated wheel connected directly to the encoder sitting on the moving part that runs along a rail or something simular?
.

As per post #11
Max.

 

Yes, like #11... Must of missed that one