Minimum constraint design

I talked to a Mech Eng friend of mine a couple of days ago about RepRap. I told him that one goal of my design was to obtain sub-millimeter accuracy (actually, I want 100um accuracy). To achieve this, I'm ok with using a bootstrapping process where the build platform goes through several iterations before coming to the target accuracy.

He brought up the concept of minimum constraint design. Here's the idea:
If you have a round hold, and you put a pin into it, you need to have precise machining or it won't work properly. Too small of a hole and you have to force the pin, too large of a hole and you end up with slop. So for the same task, the approach in minimum constraint design would be to use a v-shaped cut-out with a set-screw (or spring bearing depending on the application). This guarantees that the pin stays in the correct location, that is, it is constrained by the V cutout.

Apparently, this kind of technique has been used by several large companies to make high precision tools using low precision tools.

I'm wondering if this approach has been considered before. I don't think it creates more complexity, probably just a different thought process when laying out the design.

Minimum constraint design is what Reprap is all about. We're constantly trying to make high accuracy stuff with what are effectively stone axes.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Hi Forrest,
I understand that the goals are the same, but has the specific approach been used? I've looked at the mechanical design a bit (I admit I haven't looked at it in depth), and I've noticed quite a few pin-in-hole type joints! Are these places where constraining the pin is not necessary for accuracy, or are they places where MCD was not followed?

Considering that the Darwin is built almost exclusively on pin-in-hole joints, I think this is pretty relevant. Of particular interest are the linear slides. If we were to construct the linear slides with a V structure above the rail and a sprung skate bearing below, we could probably get a little more accuracy and a little less backlash. I'm not sure if the magnitude of these improvements will be relevant or not.

Brendan

I can't comment since I didn't design Darwin.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

I think a restriction on MCD is the available tools.

I am thinking of my mini lathe here, it has Morse Taper attachments. ie the taper makes sure that there is a no slop fit that is well centred.

Cutting tapers is not generally that easy without special tooling etc (thinking now of a tapered pin in a tapered hole).

The easy bit is turning a parallel hole into a tapered hole using a reamer or some such cutting tool (can even be driven by hand). The difficult bit AFAIK is cutting the reciprocal taper on the pin ie an outside taper as opposed to the inside taper of the hole. Particularly without access to a lathe.

The suggestion or point is a good one though, I just upgraded the deep groove ball bearings on my lathe to taper roller bearings and am impressed with the better surface finish I now get.

Somewhere I suggested putting taper bearings was on the Z axis ends.

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Pressure fits are an acceptable solution for backlash removal.
Where a joint is meant to have movement or slide though, creating pressure through additional hardware as Annirak mentions, is sometimes needed. Less backlash usually means more driving power is required though.

Ah! Now I see why Annirak needs the more powerful stepper motor drivers !