Metric or UNC Threads in Aluminium

[William Harvey 1Participant @williamharvey1]

Hi,

I need to secure a steel trigger wheel to an aluminium crank pulley.

The pulley and trigger wheel will be bolted to the crank (already sorted that bit) but I needed a way of ensuring that the trigger wheel does not move around its centre in relation to the pulley, otherwise the engine timing will be thrown out.

So I thought at first a dowel pin, but for added securing I thought I'd rather put in a screw.

The pulley outer ring is about 15mm.

I am thinking a couple of M4 button head allen socket bolts, but will a metric thread be OK? The idea is that I will drill through the trigger will 4mm and then drill and tap the pulley below for an M4 bolt about 6mm deep.

So first up what are the thoughts on the above plan?

Secondly – I got confused and thought I would need a UNC threaded bolt as on these engines bolts in aluminium are UNC. A colleague suggested 3/16 UNC, but I am having problems finding the correct tap and drill for 3/16 UNC.

The trigger wheel is 5mm thick.

In the pic above most of the bolts that secure the trigger wheel to it's centre boss have been removed temporarily.

[William Harvey 1Participant @williamharvey1]

[JasonBModerator @jasonb]

You won't find it described as 3/16 UNC as below 1/4" they change to a number system so that would be #10-24 where the 10 is approx 3/16 and the 24 is the tpi but that is closer to M5. if you want M4 size then #8-32 would be closer.

Now 32 tpi is approx 0.8mm pitch so the M4 with it's 0.7mm pitch will actually be coarser. On the other hand 24tpi is coarser than the 0.8mm pitch of M5.

On a shallow blind hole you may be better off thinking of number of turns of engagement rather than just the coarseness of the thread

For the likely load I doubt it makes any difference.

Edited By JasonB on 06/07/2022 08:09:53

[Kiwi BlokeParticipant @kiwibloke62605]

ISO metric coarse series fixings are generally OK in Al alloys, as are UNC. However, there's not much point using UNC these days, since metric fastenings and tooling are more readily available, and it's the modern standard.

[SillyOldDufferModerator @sillyoldduffer]

As a general rule fine threads are stronger than coarse and less likely to vibrate loose. But for most ordinary purposes coarse threads are preferred because they're faster to fit – easier to align, less chance of cross-threading, and able to cope with a bit of dirt and rough cut threads.

There are always exceptions! Thin walled pipe, soft or brittle materials, high-vibration etc. Aluminium might be an exception except I don't think UNC offers any advantage over Metric Coarse; apart from being dimensionally incompatible, their mechanical performance is similar. Whitworth is probably best choice for soft Aluminium because it's form was derived from industry experience fastening early 19th century cast-iron.

In this application, the bolts aren't stressed in a way that would cause threads to strip. However, as the assembly is part of an engine, vibration is likely to be a problem and Whitworth is also the thread form most likely to vibrate loose.

I'd use Metric, simply because it's easier to source in the UK and usually cheaper. North Americans are more likely to choose UNC bolts for the same reason – availability. And whatever thread was used, I'd use a strong high-temperature thread-locker.

Dave

[HopperParticipant @hopper]

Countersunk screws could give you more positive location, not allowing the possibility of the wheel moving by the amount of clearance between the holes and the screws at high rpm under intermittent loading as in an engine. They perform the function of a retaining bolt plus locating dowel in one.

+1 on use coarse series threads in aluminium. Bit of Loctite blue threadlocker on fasteners whirling around like that is always good too.

Edited By Hopper on 06/07/2022 10:38:58

[Kiwi BlokeParticipant @kiwibloke62605]

I thought it was considered bad practice to load countersunk-head screws in single shear. In this application, a dowel or two would be appropriate: it sounds like the OP has done that (I think…).

[HOWARDTParticipant @howardt]

Including a couple of dowels, or one if the parts are spigoted would give a positive drive. What ever screws you use there will be clearance holes to allow movement between the parts. Thread type doesn't matter, you could always use helical inserts.

[duncan webster 1Participant @duncanwebster1]

Coarse threads are stronger in non ductile materials, eg cast iron. When I worked in aerospace all tapped holes on ally were helicoiled from new

[Ian PParticipant @ianp]

Whatever fixings are used to hold this disk to its hub, holding it to the hub is their only function.

I cannot see how there will be any shear forces involved. The disk will be subject to an insignificant amount shear force caused by angular velocity changes with engine RPM it will also have some wind resistance created by the notches in the periphery. The only reason for dowels of other locating methods might be to ensure timing repeatability (say after reassembly) but otherwise holding the disk in place with three small loctited screws is all thats really needed.

Just as an observation. The disk appears to be laser cut but looks more like SS than carbon steel, so I wonder if this for an optical sensor?

Ian P

[William Harvey 1Participant @williamharvey1]

Posted by Ian P on 06/07/2022 14:42:41:

Whatever fixings are used to hold this disk to its hub, holding it to the hub is their only function.

I cannot see how there will be any shear forces involved. The disk will be subject to an insignificant amount shear force caused by angular velocity changes with engine RPM it will also have some wind resistance created by the notches in the periphery. The only reason for dowels of other locating methods might be to ensure timing repeatability (say after reassembly) but otherwise holding the disk in place with three small loctited screws is all thats really needed.

Just as an observation. The disk appears to be laser cut but looks more like SS than carbon steel, so I wonder if this for an optical sensor?

Ian P

The sensor is magnetic I believe?

[Pete RimmerParticipant @peterimmer30576]

More likely an inductive sensor. Either way for your application a standard metric fastener (or several) will be fine. Personally I would use three, and I would slot the timing wheel holes for adjustment, then when it was all timed in I would loctite them in.

[Howard LewisParticipant @howardlewis46836]

Normally in Ali you would go for coarse thread, to have a deep thread, particularly for consistence with the rest of the engijne

But the trigger wheel will see not torque loading, apart from acceleration / deceleration torques.

The clamping forces should provide sufficient grip to prevent relative movement, but a dowel would make certain.

I would be inclined to take Jason's advice and use 10-24 UNC.

Howard

[Pete RimmerParticipant @peterimmer30576]

Posted by Howard Lewis on 06/07/2022 22:30:25:

Normally in Ali you would go for coarse thread, to have a deep thread, particularly for consistence with the rest of the engijne

Howard

Japanese ally-block engines are riddled with metric fine and metric extra-fine threads. My own bike used M12 x 1.25mm head and main bearing bolts that were torqued to 75 foot pounds.

[Neil LickfoldParticipant @neillickfold44316]

If the alloy is soft, use thread inserts, like 2d helicoils and then just use your normal M4 or M5 threads. In magnesium castings we used helicoils from new. In harder alloys like 7075, then you can just use metric threads with no issues. In the softer extrusions if there is any chance of it moving radially, add some roll pins to help it lock into place.

[Author]

Posts