How to specify dowel pin locations

[kego]

I want to specify dowel pins for a clearance fit. From my understanding I need an H6 fit. I want a fit that provides a true alignment of two machined pieces but is able to be taken on and off by hand.

I have seen a lot of information on the tolerance of the holes but i have not seen anything on tolerancing the position of the hole and this is the area I am troubled.

 

[powerhound]

Make one clearance hole an H6 fit and the other a slot with the width toleranced to H6.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[TheTick]

Typical dowel pin locating schemes will use one hole and one slot on the mating part.

 

[semiond]

powerhound, TheTick,
I don't know if the OP got his question answered or not but i'm intrigued.
Do you mean a scheme where there are 2 pins fitted in holes on one part, and one slot on the mating part inside which both pins should fit?
This is the only way i can imagine such locating scheme could be robust and yet not overconstrained.

Sorry if i'm asking something obvious. I'm inexperienced in this kind of applications.

 

[rothers]

2 pins fitted in holes on one part, one hole and one slot in the mating part, length of slot is aligned with pin centres.

 

[semiond]

rothers,
The scheme you describe constrains another degree of freedom in addition to what i had in mind (translation along the slot).
But this requires accurate positioning of the slot's center plane in relation to the hole axis in the same part, and i think the slot in that case should be slightly wider than H6 edit: at it's MMC condition than the maximum hole diameter size to compensate for the position deviation. Am i right?

 

[PaulJackson]

Kego,
I don't know what type of joint you are designing but there are many designs that incorporate two round dowels and two round dowel holes. Almost every coupling where axes on either side of a housing joint must align (where alignment is not provided by the components alone) use two round dowels and two round holes i.e. engines & transmissions, motors & devices. The spread between the dowel pins and dowel holes (although identical dimensioned) are rarely equivalent so the pins are either snug on the near sides of the holes or snug on the far sides. The coupling makes both dowels responsible for rotation & translation where with a hole and slot one dowel stops translation and both stop rotation.

Since the pair stops both translation and rotation it typically constrains all remaining degrees of freedom not constrained by the joint face there is no need for a tertiary datum feature. The dowel joint is typically designed with slight interference either inboard or out of the spread and the solid pins have generous lead in radii or chamfers but split hollow or rolled spring pins are also used. For the GD&T both pins and holes are specified as a 2X pattern with a position tolerance RFS to the structure datum features then the spread is controlled with a position tolerance MMC to the joint face only. Naturally the basic spread is identical for the holes and slots and you could use position O@MMC to determine your min hole, max pin but then you must work out max hole and min pin accounting for fit, mfg tolerance, and structure rigidity.
Paul

 

[drawoh]

kego,

You have correctly identified the problem. If you have two pieces located by two sets of accurate holes, you have to account for the positional tolerances of the holes, preferably without super-duper accuracy. There are two ways...

[ol 1]
[li]You slot one of the clearance holes.[/li]
[li]You use a diamond shaped dowel pin[/li]
[/ol]

Either way, the distance between the dowels and the clearance holes is not critical. Note how the first hole locates in X and Y, and the second hole controls rotation only.

--
JHG

 

[kego]

Thanks everyone.

Does anyone know a good video series or book to brush up on my GD&T?

 

[drawoh]

kego,

ASME Y14.5-2009. You might as well go to the source.

--
JHG

 

[semiond]

I agree with drawoh,
That's the only GD&T "book" i ever used.

 

[powerhound]

I agree that the standard is a great resource but I don't think it's a very good textbook to learn from. I'm sure some of your colleagues have textbooks from courses they've taken in the past.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[semiond]

Maybe i'm a little late in continuing this discussion, but does anyone else share my observation that making a hole and a slot on one mating part doesn't provide a great advantage in terms of ability to loosen tolerances?

The distance between the hole and the start/end of the slot will no longer be critical, but the alignment between the center plane of the slot and the axis of the hole still must be very accurate. I think that the maximum material condition of the slot in such application must be srtictly controlled for position in reference to the datum hole. On the other hand if a bonus is allowed, one may end up with a wider slot than needed for a rigid (location edit) orientation of the part.

 

[chez311]

semiond - the use of a slot doesn't mean that the slots location/orientation/size isn't still critical, (though I would say orientation is less critical than the rest - imagine a slot which is rotated about the center axis of a mating pin, additional rotation *note PURE rotation* of the slot does not result in significant additional misalignment of the pin) I would say the main advantages of utilizing a slot would thebe for assembly. If 2x holes and 2x round pins are utilized production at/near the edges of the tolerance bounds can result in binding and issues during assembly, or even a condition where assembly is not possible even if technically it is within tolerance (ie: when any slight misalignment results in binding - especially with heavy parts). Such issues are significantly reduced utilizing a hole and slot and most deviations simply result in misalignment instead of misalignment and binding.

That being said, as PaulJackson pointed out there are many applications especially in the automotive world that utilize 2x round pins and 2x round holes with very few issues with assembly without unreasonably tight tolerances. One of the key items here is probably keeping the mating pin length as short as possible.

Edit: grammar

 

[3DDave]

One factor not mentioned is that large parts can deform small amounts with low levels of installation force. Between deformation from interference fits to other elastic movement, the tolerances required are dependent on the amount of force that can be used to assemble/service the item. Most tolerancing is described in terms of clearance fits for drop-in assembly techniques so the forces and deformations are assumed to be zero, but that is often not the case. For example, the reason for alternate tightening of fasteners on such things as engine heads and wheel lug-nuts is deformation.

Misalignment with a slot is typically a negligible error source if the misalignment is less than about 5 degrees due to the sine component being so small. As the mating part shifts the distance between the pins that is multiplied by the sine of the angle to result in side-to-side displacement. Since slots are typically not interference fits, the amount of side-to-side displacement is often on the same order of magnitude as the tightest slip fits. In any case, it is easy, and necessary, to do the calculations regarding all contributors.

 

[drawoh]

semiond,

If I fabricate one piece with two dowels and the other piece with a hole and slot, I have an assembly that is repeatable, not accurate. We don't need the slot to be precisely lined up with the hole. If you need as-assembled accuracy, you have a far more challenging design problem.

In general, dowel clearance holes are accurate enough that there is no point fiddling with MMC/B.

--
JHG

 

[semiond]

Chez311, 3DDave, drawoh - Thank you for the contributions. Very informative and sheds more light on the subject.