Member postings for Andrew Johnston

I wouldn't know, never contemplated building any sort of Stuart model.

Andrew

Even if the sides are supported by the tapped hole? Of course that means the grub screw will be completely within the tapped hole, but isn't that how a grub screw is supposed to be?

Andrew

While the propensity of a boring bar to chatter is related to the unsupported length it is less directly related to the deflection. Chatter is not caused by deflection, it is a cyclic change of deflection that causes the problem. It is counter-intuitive, but a way to kill chatter when boring can be to increase the depth of cut and/or increase the feedrate. This may increase the deflection, but makes the deflection stable; so no chatter.

Andrew

Among other things I finally got the overarm on the horizontal mill shifted from this:

To this:

It maInly involved BF and BI, ie, hitting it with a copper hammer. Once I'd got the arm moved 8" or so, plus copious amounts of oil on the dovetails, I could move the overarm all the way in and out simply by pushing on it. Which is where I was 10 years ago. Must remember to exercise it now and again and add oil. The reason for moving the overarm back is so I can use the face mill shown to start roughing out the cylinder castings on my traction engine without said overarm getting in the way. The face mill may look small, but it is 80mm diameter.

I might even get around to fitting, and trying out, the vertical head for the mill I bought years ago on Ebay:

That's a 12" rule against the head. The main issue is working out how I'm going to lift the vertical head onto the mill table without involving a trip to A&E. When I bought the head I lifted it from the car boot to where it is now. But that was years ago, no way I can just pick it up and lift it now.

Andrew

Indeed, but it's still pretty fast. It equates to 189fpm, which is close on double the recommended speed for HSS in mild steel. I reckon there are 22 teeth on the cutter. So at 4 thou per tooth times 120rpm that equates to a feedrate of 10.56 inches per minute. If we assume the table feedscrew is 8tpi that's about 0.7 seconds per turn of the handle. That's going some!

From the video the belt looks pretty tight, so I think I'll modify my advice and say too big a cutter and/or running too fast and/or too slow a feedrate.

Andrew

Could be; depends upon cost, cutter type, condition, bore size and location. I'd be particularly interested in slab mills.

Andrew

On the assumption that the machine didn't vibrate when new the question is why is it doing so now? Poly V-belts may be better, but they're a red herring while the source of the vibration is unknown. A few ideas are:

A 10mm deflection for 10N (1kgf) force seems pretty slack to me, especially on a long unsupported belt. None of my V-belt driven machine tools have a vibration issue, and none of them have idler pulleys.

The second belt looks a bit odd; is the top pulley in line with the intermediate pulley or should the belt be one pulley further out at the top?

While it's unlikely to be the problem a single phase motor is not as smooth as a 3-phase one. It might be that the pulsing characteristic of the single phase motor is exciting something it shouldn't.

Would we be correct in assuming that the vibration only happens when a cut is in progress? If so what is the feedrate? A good starting point for chip load is 4 thou per tooth. When I started using my horizontal mill I used a much smaller chip load, and the whole machine vibrated. My horizontal mill weighs nearty two tons.

What rpm is the cutter running? Looking at the speed chart the lowest value of 150rpm is on the fast side for a 5" cutter in steel.

Andrew

How far can you move the V-belt at the centre of the unsupported length? Is the replacement motor single phase or 3-phase?

Andrew

Rod: At least you got a file out of the system. I need to do some printing for work. So this morning I downloaded the latest Cura (4.?) on the new computer. Ducking useless! All it did was whinge about a disc not being in some obscure imaginary drive. It never started properly and I couldn't get rid of the annoying pop up about the drive. Had to restart the computer in the end. The "new" versions of Cura did this a few months ago when I last tried updating. So I reverted to 15.? on the old computer. That didn't seem to display brims properly. So I gave up and used the g-code file I already had.

Sadly it seems that Ultimaker have gone the way of many start ups in believing their own hype. So we get over-complicated software with unwanted "features" that plain don't work. Worse than a chocolate teapot - at least you can eat the teapot.

Andrew

Sorry about that, I won't do it again, honest.

Andrew

No it isn't, the thread depth for a 4tpi Whitworth thread is 0.16008" from the radius on the crest to the radius on the root. So a sharp pointed tool plunged by the thread depth will not cut the proper thread profile.

Incidentally 2" BSW is 4.5tpi, not 4tpi.

Andrew

I tend to use much larger sections, so may be they're less prone to distortion. When I made some press tooling for the flypress I skimmed the surfaces of hot rolled steel to form the blanks:

After the curved surfaces were CNC milled the tooling fitted together very well; certainly less than a thou or two gap:

But these blocks are about 5" x 2.5" x 3/4".

Andrew

That's interesting; I've machined a lot of hot rolled steel and never noticed a problem with distortion. How did the problem manifest itself?

Andrew

A fixture and a custom reamer is the way to go. Do the drawings call out a tolerance and surface finish value? I don't suppose the aircraft manufacturer can be identified, but it would be useful to know. So I can avoid flying on them.

Andrew

I made my cam from a solid piece of gauge plate. But I bought my Economy kit many years ago and didn't get much in the way of materials other than the castings. I did get a flame cut conrod (eventually) but binned it as it was too thin. I also binned a number of the castings, especially the smaller light alloy ones.

There are, or where, a number of errors in the drawings. One blooper is the requirement for 1/2" BSP threads for the exhaust and carb fittings. The designer obviously assumed that 1/2" BSP was 1/2" OD. I used M12 instead.

Andrew

The head on my Bridgeport mill tilts in two axes. It's not a feature I've used more than a handful of times in the last 20 odd years. I prefer to get the required angles with an adjustable angle plate or other fixtures. I believe that later Bridgeports had dowels to reset the head to vertical, but if I move the head i have to tram it afterwards. I tend to check the tram on a regular basis as the head can move under heavy cuts. So definitely an area where rigidity is lost in exchange for potential flexibility (pun intended).

On a practical note the Bridgeport head is very heavy, and even with the worm drives it is hard work moving it and then retramming.

In theory the head can be moved 90 degrees to create a poor mans horizontal borer, but when I need such a facility I tend to use the horixontal mill. For drilling and light milling parallel to the table I prefer to use a right angle drive rather than move the head.

Andrew

Edited By Andrew Johnston on 26/10/2019 10:33:24

I'm not sure what is meant by an adjustable hand reamer? Something like this:

This type of adjustable reamer is intended to take out minimal material per cut, mainly for repairs or fitting oversize pins in the field. They're not easy to use or particularly precise, and as Hopper says the hole in question is really too short.

Without knowing the application the design sounds a bit of a dogs breakfast. We use plenty of push fit bearings in aircraft and they're all staked into steel housings, so I'm puzzled as to the use of aluminium. Personally I'd corner the designer and quiz him on the design and machining instructions.

Andrew

You can design in any scale you want and then scale to final size in CAD. Having said that I normally design to the finished size of the part. An exception is gears that are going to be machined on the CNC mill. In that case I design as 1DP and then scale accordingly. Personally I do any such scaling in the CAD system. When scaling a 3D print for shrinkage I use the slicing software for the 3D printer. One problem is that the shrinkage of the printed part is anisotropic and also varies with part size, shape and layer thickness. I've recently 3D printed a trial cylinder flange for my traction engine to check fit on the boiler:

I scaled this by 0.5%, which turned out to be generous, as the length of the flange is about 0.2% too long. I think that the shrinkage of parts created from liquid resin is smaller than parts printed from filament, if only because heat isn't involved.

I think you'll find that shrinkage in a casting is more complex than a scalar value. It will depend upon how the casting cools. Once part of the casting has cooled it isn't going to shrink any more, even if other parts of the casting around it are still shrinking - look up hot tears. The shrinkage will also depend upon any cores and how they fit within the casting.

Of course it all depends what sort of accuracy you hope to achieve, but it isn't always simply a case of applying an overall scale factor. For reference the flange above was about 14 thou oversize in 7 inches.

Andrew

Must remember not to be a smarty pants.

I will admit to breaking an 8BA tap in gauge plate recently. But that was because I was clumsy, nowt to do with the tapping drill size. The tap broke well above the hole, and the broken piece was loose and unscrewed with fingers. The hole was the last of 8 and I got careless and applied a bending force, equals a snapped tap.

Andrew

In ordinary steel I'd drill 7 to 7.1mm. In brass or aluminium 6.9 to 7mm and in stainless steel 7.3mm. In percentage terms I aim for 60% to 70% in most materials, may be 75% in softer materials and 50% in tough materials. The foregoing applies to coarse threads, for fine pitch i'll go down a tad in size.

The drill size has little or no effect on the fit of the thread. That's largely determined by the tolerances of the tap and mating screw.

While not universal the bolt usually fails in tension long before the internal thread fails in shear. Many years ago I did some experiments tapping 6082 and using high tensile (12.9) bolts. On the first test the bolt broke before the thread stripped and that hole was drilled for 50% thread engagement.

Andrew