4
$\begingroup$

I remember hearing awhile back that it's not a good idea to use screws to locate/align parts. Why is that? Apparently dowel pins, or "locating pins" should be used instead. From my understanding, I suppose it is because there are much tighter tolerances when manufacturing those pins compared to fasteners.

Does this inherently mean that you should not use dowel/locating pins in a design when you are screwing into a threaded hole? Since there might be binding. In addition, if you were to decide you wanted to use locating pins/dowel pins for a more precise connection, then your holes for your fasteners should be not threaded (thru holes) and sized a bit larger than the shank diameter to allow some movement.

$\endgroup$
3
  • 1
    $\begingroup$ "then your holes for your fasteners should be not threaded (thru holes) and sized a bit larger than the shank diameter to allow some movement." When fastening two parts together, the hole on the part where the screw enters should never be threaded. Ever. If it is. it'll bind because the thread is not continuous from one part to the other unless you tapped both parts while they were clamped together. And if you did that, it's not going to stay that way since the first thread always tears out. $\endgroup$
    – DKNguyen
    Oct 16 at 4:47
  • 2
    $\begingroup$ which is why holes are counter-drilled or counter-sunk slightly, to give that torn thread room to sit so it doesn't stick up like a burr and stop the parts from sitting flush against each other. $\endgroup$
    – DKNguyen
    Oct 16 at 4:50
  • $\begingroup$ The manufacturing tolerances for threaded fasteners are just not up to snuff, and they wear when you assemble and disassemble stuff. Pins are entirely inside and protected. A hydraulic pump might need 12mm bolts to hold it together during operation, but can be assembled with 0.5mm pins on the bench. $\endgroup$
    – Phil Sweet
    Oct 18 at 1:48
3
$\begingroup$

You seem to have most of it.

Take a waterpump on an engine: dowels are used for precise location as the impellor needs to be located with the back of the pump on the engine.

Some items use bolts with threads and a shank to give a controlled location.

Other parts such as wheels use a concentric flange on the hub with a matching hole on the wheel so that the bolts or studs only provide a clamping force and the studs or bolts have a non threaded portion, the shank, to protect the holes. Also the bolts or nuts have cones to locate the wheel with matching ones on the wheels.

$\endgroup$
3
  • $\begingroup$ I see, thanks for the clarification! Much appreciated. Just a question on "Some items use bolts with threads and a shank to give a controlled location.". Do you have an example of this? $\endgroup$ Oct 16 at 3:11
  • $\begingroup$ Fish plates joining beams together, the shanks have a given size, the holes are drilled to a tolerance and the beam gets licated. Washer and nuts make sure the nut does not run out of thread on the bolt, part of the specification… $\endgroup$
    – Solar Mike
    Oct 16 at 3:15
  • $\begingroup$ @TwoWaySpeedOfLight Here's a shanked bolt in situ: i.stack.imgur.com/iQNEx.png => you could think of it(if the tolerances were really, really close between the hole and the bolt) like a "bolt with built in locator dowel". You can find examples of bolts with shanks larger than the threads too, but it doesn't really make sense to have them smaller. The tips of the threads in that pic are the same diameter as the shank, but it's not great to use such for locating because it's essentially a saw and any movement will cause the bolt threads to "saw" or "grind" at the sides of the hole $\endgroup$
    – Caius Jard
    Oct 18 at 8:43
5
$\begingroup$

Screws and their holes have a few of problems:

  • Screws are formed, so nowhere near as accurate as a hardened, ground pin. You probably could harden then grind a screw thread but obviously much more expensive. Probably only used on things like micrometers.

  • The tapping operation itself removes a fair amount of material. Second only to the drilling operation. Large cuts require large forces which produces lots of heat which results in heat expansion, deflection of the tool, and compression (and resulting rebound) of the material, all of which reduce accuracy of the hole.

  • For applications where location really matters, you can't even bring the parts together with screws because the tolerances are too tight, even if you had super accurately ground screws and super accurate tapped holes:

    If you have to do something like placing the top half of a heavy steel die onto the bottom half you can't really do that with locating screws installed; They'll just sit on the screw tips. So now you have to rotate all the screws at the same rate or else it will tilt and jam and not only that you have to somehow position each screw relative to it's hole before you start turning
    all at the same speed because they must all engage their threads at the same time.

    Nor can you place the top die onto the bottom die and then insert screws because the part has nothing to locate against while lowering. What are you doing to do? Set the parts against each other and then bang it over and over again with a mallet as you try to adjust the position so that you can get the screws to go through? *Actually, pins have a similar but solvable problem here too where if the parts aren't lowered exactly right the pins will jam and the part won't go down, even if the pins are chamfered. It will catch and bind on the chamfer. Apparently the solution is to add a small ring of rounded undercut or relief right behind the head of the pin and radius the head so it curves into the relief. Then you radius the hole that it mates with. If you size everything correctly, the half being lowered can catch and rotate within the space provided by the relief as it comes down and self-locate. I don't know the name of this but it's been around a very long time.

Since the pin is more accurate you can also make use of a more accurate hole. To make a hole for a pin you first drill the hole (to make a hole, duh) but the drills are designed to remove material and remove material fast so the hole doesn't have a very good finish and isn't as centered, straight, or round. They hog material and can flex and dig their own path through the material.

So you then bore the hole which shaves off the periphery of the hole and since the tool itself spinning around a location centered in the air (in the hole made by the drill), it ensures the center of the hole is where it should be be. Boring makes sure that holes are located where they are supposed to be.

Finally, you ream the hole shaves of small amounts of material off the edge to ensure that the hole is the diameter it should be. Reaming ensures that holes are straight and of the proper diameter.

Skip reaming and your hole may be centered at the correct location but be the wrong diameter. Skip boring the hole and your hole may be straight but off-center (from where you need it to be). Not that you couldn't do those operations before tapping the hole, but those are some crazy excessive operations for most screws.

$\endgroup$
3
$\begingroup$

IMHO, the main reason is that using multiple screws for locating the parts introduces additional stresses on the bolts and makes them susceptible to failure. So especially in high stress environment, cyclic loading or high temperatures, they are usually avoided.

So problems start to rise when there are misalignments. Like the following.

enter image description here

Things are worse in 2D (in blue are the holes of the bottom plate, and with red on the top plate)

enter image description here

Having said that, there are screws that can be used for locating parts e.g.

Name image
Locating screws enter image description here
countersunk bolts enter image description here
$\endgroup$
0
$\begingroup$

Our policy for assembling modest precision electronic instrument chassis was to have one countersunk screw that would locate two parts, and any other screws retaining them together would have a head that was flat underneath, to allow a tolerance in hole centres to accommodated.

$\endgroup$
5
  • $\begingroup$ This is only reliable if you fit the screws in the correct sequence. Otherwise, someone will assemble and tighten the flat head screws one by one, and then leave out the countersink screw because "it doesn't fit properly." (Or they use more force and just re-cut the holes and threads!) $\endgroup$
    – alephzero
    Oct 17 at 11:16
  • 1
    $\begingroup$ @alephzero Ohhh! You've worked with 'Other People' (tm) before! $\endgroup$
    – Neil_UK
    Oct 17 at 11:30
  • $\begingroup$ @alephzero What are the two of you referring to when you say "flat head screw" because where I'm from flat head screws are flat on top with taper and need to be countersunk. Do you mean pan head screws? $\endgroup$
    – DKNguyen
    Oct 17 at 16:36
  • $\begingroup$ @DKNguyen Where I'm from, if someone says countersunk screw, then in the same sentence says flat-head, it's obvious that the distinction is with the non-flat part of the previously mentioned countersunk screw, regardless of whether the screw is the type known in some parts of the world as pan-head (rounded top, flat bottom) or cheese head (flat top, flat bottom), mushroom, fillister, button, truss or dome head, all of which would also tolerate misaligned holes. But I guess I need to update my answer, $\endgroup$
    – Neil_UK
    Oct 17 at 18:19
  • $\begingroup$ @DKNguyen What Neil_UK said. (Yes I admit my terminology was a bit careless.) $\endgroup$
    – alephzero
    Oct 17 at 18:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.