Methods to reduce assembly gap

[Jonathan Tang]

Hi guys,

Let's say we have the assembly of three blocks placed in a container such that the worse case dimensions of the block and the inner width of the container results in zero gap. Therefore, in non-worse case situations, there will be a non-zero gap on the right side.

Question is: What methods can we use to reduce this gap further?
Any assumption on type of fit, block and container material etc is allowed.

For example, two methods are 1) Use RSS so that nominal gap can be reduce. 2) Built in interference but cool the blocks so it shrinks and heat up the container so it expands so that the blocks can fit.

I can't think of any other methods - a press fit? But not sure how that would work.

 

[3DDave]

RSS barely works if the distributions of the widths are all Gaussian and still provides a Gaussian distribution anyway, not a firm cutoff.

Peel-able or stacking shims are a popular choice. Start with a stack for the worst case gap.

Blocks of what?

 

[Jonathan Tang]

The blocks can be any material - does that answer your question?

Shims are a good idea.

 

[3DDave]

No, it doesn't, but different materials have different failure modes so if those happen ask more.

 

[Scuka]

First of all, there's important info missing here. Without that, we're just shooting in the dark.

What's the purpose of reducing the gap?
What's the size of blocks and containers?
What's the range of acceptable values for the amount of gap? Is it 0 (effectively press fit) or some non-zero amount?
The size questions are crucial because obviously fitting cars into a shipping container is very different from fitting 1 cm machined blocks into a machined container.

But some of the possible solutions:

Moving sidewall that is adjusted after placing the blocks in.
Compliant sidewall.
Compliant fourth block.
Fourth block manufactured after the gap is measured.
Tapered fourth block.
Stacked shims.

 

[SirPhobos]

I think the most basic answer would be to tighten up your tolerances.

If the container length is 3 -0/+.1, and each block is 1 -.1/+0, then the largest gap is 0.4.

If you tighten all those up by half, the container becomes 3 -0/+.05 and each block being 1 -.05/+0. This results in a maximum gap of 0.2.

-- SirPhobos

 

[geesaman.d]

Ripple springs.

Statistical tolerancing.

 

[Jboggs]

This question has all the earmarks of a theoretical classroom exercise. Doesn't sound like a real world problem at all. Is it? If so, what are the real parameters and limitations?

 

[btrueblood]

Pourable (or squirt-able) urethane foam. Has the advantage of also gluing all the blocks and the container into one solid mass.

 

[TheTick]

Whoever is giving you grief about this needs to accept that the gap is the result of the processes. It is not the result of "lack of vision" or "failure to think outside the box" (pun not intended but still enjoyed ).

The problem is indeed as simple as it appears.

Depending on the scale, opposing wedges or threaded posts may serve to prevent movement.

If you're going to achieve fit by press or thermal means, do not underestimate the amount of stress.

 

[Tmoose]

The OP question is too open ended.
Fit versus function.
Is zero gap the goal, or precisely as stated, simply some amount of gap reduction ?
As mentioned by others, The effects of thermal expansion in service could be mighty significant, so are among the details that really need to be included in the initial problem definition.

https://www.henkel-adhesives.com/us/en/products/industrial-adhesives/retaining-compounds.html

How about making the opening in the housing real big, to include one of these?

https://cdn.mscdirect.com/global/images/ProductImages/8640717-21.jpg

Or just adding some tapped holes to one end of the main housing for a few setscrews to clamp the parts securely, or even to just maintain .005" "gap" ?

yeah, the OP question is too open ended.

 

[BrianE22]

What does "RSS" stand for? Measure and sort at assembly? If so I believe that is what's done with the balls in ball bearings

 

[SwinnyGG]

RSS = Root Sum Square tolerance analysis. It gives you a statistical estimate of combinations of parts, but requires very specific statistical distributions within the real tolerance window in order to work well.

This reads like a homework problem for a 100 level engineering class...

 

[rQuestionEngineering]

Get to know Non-Shrink Grout - a very useful material
Non-shrink grout is a cement material with many uses. Unfortunately, most big box stores hide it away and direct you to the ceramic tile section if you ask for it.

It is generally hidden in the masonry (concrete) section.

Non-shrink grout is usually rated 8,000 psi, so it is very strong stuff. The cool thing is you can mix it from near dry (like clay) to heavy cream consistency. The "non-shrink" part means that you can mix it like heavy cream and pour it and whatever size or shape it had wet it'll have once it hardens.

For example, I wanted to install a drain in my wife's greenhouse so I cored a 3" hole and ran a 2" pipe through it. So I stuffed some foam down the hole an mixed up some wet grout and poured it in. This is what it looked like before and after it hardened

http://i.imgur.com/thxCodE.jpg

Now, you may not want to fill a hole in concrete, but say you have a beam or something you want leveled. You could install shims or something, but most of the time all the weight is on a very small part of the shim.

I had to install and level my car lift. Rather than using shims permanently, I shimmed it level with and built a little dam about the base. Mixed up some grout which I poured near the base, the grout flowed underneath the posts providing a perfect 8,000 psi level base.

Grout is your friend!