andypugh

Productivity enhancement. I added a stripper and a foot-operated air blast to clear the completed parts. 

https://youtu.be/Bpb68xB6zTc

8 parts in 45 seconds, down to 20 hours to make the set. 

I had the idea last night that I should have drilled the lower follower for air blast, and then could have used the piston action to valve an air blast from the pressurised lower section. But it's all hardened now, so too late to be drilling holes. 

Soldering is likely to be fun. I would imagine that assembly with solder paste and then running over each tube with a torch would be the cheap way. (melting enough solder to dip a complete tube would be the quickest and easiest). 

I have experimented and found that the fins sucept well in an induction heater coil, which would be a high-tech way to melt solder paste. 

1 hour ago, Old Bill said:

Nice job Andy. Is that one action or two?

One action to make the square, torus and petals. But no petal-turning-over. 

https://youtu.be/j0FX1ER4URY

In the video above I make 5 in a minute. So 14,000 would be 46 hours of work. 

It could be much faster with a stripper to pull the sheet off the punch, and with an air-blast (maybe on a foot switch) to kick the finished part into a hopper. 

On 12/22/2020 at 6:03 PM, Asciidv said:

Andy, Ben, if you want sharper corners I can wire EDM the bits. It might be an advantage to wire EDM everything from tool steel so that hardening is not necessary,

I think that I could slightly re-design the existing punch and die so that EDM-ed pieces would cut the profile, while keeping hardened silver steel for the parts that do the forming. 
Probably an idea to keep in reserve if tool life proves to be an issue. 

Would these do? 

10 minutes ago, Asciidv said:

I like this punch. Simple to make and producing 40 in a single stamp. 

Looking at the eccentric layout of the punches I think that these probably stayed in one piece (which would make for a stronger but impossible to repair radiator)

2 minutes ago, Asciidv said:

Andy, Ben, if you want sharper corners I can wire EDM the bits. It might be an advantage to wire EDM everything from tool steel so that hardening is not necessary

It would also be an easy way to get back-relief, 

Though my current design has turned features on the punch and follower. (the earlier link should now show a rather re-worked design, using much smaller bits of tool steel. 

19 minutes ago, BenHawkins said:

Is bending the tips of the petals as simple as forming a flared hole before starting to cut?

Looks tricky to make, how about you try it 🙂

I have been slightly busy...

14 minutes ago, andypugh said:

And bigger is easier, as I don't have a slotting head...

With the latest update, this is less of an issue as I abandoned the square follower design, so the square die recess is only 12mm rather than 30mm deep. This makes it rather easier to mill with a 4mm cutter. 

6 minutes ago, BenHawkins said:

3/32 but as you spotted I think the original was less.

And bigger is easier, as I don't have a slotting head...

I have what I think might be a workable (and makeable) design. 

https://a360.co/3pagTbb

7 hours ago, Asciidv said:

I wondered about a heavy duty spring although in reality I think it will have to be a latch.

https://en.wikipedia.org/wiki/Types_of_press_tools says: 

In general, cutting force (CF) can be calculated using the formula: CF =L x S x ζmax

The square has approx 90mm perimeter, is 0.25mm thick and the shear strength is about 275N/mm2:

So to blank out the square needs about 6kN. 

Zoro list a 50mm dia die spring with a stiffness of 700N/mm so even that would be compressed by 8.5mm. 

Cutting force could probably be reduced by faceting the die, as is done with Q-max cutters. 

5 minutes ago, BenHawkins said:

Thanks for everyone's effort on this. Here is a photo showing the recess on the profile from the underside on the "new gills"

What is the corner radius? (not that there necessarily is a corner radius on the original ones) 

The torus is easy, I think. 

2 minutes ago, Asciidv said:

O.K. Andy, let us see you turn the tips over!

I have given that some thought, and I think it has to be a second-op, I can't see a way to get the tip-turned shape out of a die that can also make the petals. 

21 hours ago, andypugh said:

And would a hexagon work?

Each facet of a hexagon would make 1/6 of a gill.....

It seems that the faceted punch works just fine, which simplifies things:

4 hours ago, Asciidv said:

Andy, I think the simple pyramid would just tear the metal as it forced it’s way through without any cutting action. 

It might need to be pointier. 

I think that, ideally it would be a double-action, where the square and inner ring is punched first,, then the spike comes up through the clamped, shape-locked ring and bursts it. 

But I don't know how you would do that. Possibly with a spring, but a spring stiff enough to punch the initial hole?

49 minutes ago, ruxy said:

Such as  Kendal & Gent  'chasers'  were a common cheat ,,

https://www.ebay.co.uk/itm/Rare-Vintage-Kendall-Gent-Screwing-Machine-Chasers-Tin/372646617948

Ironically the tin with contents would probably have listed at a lower price. 

37 minutes ago, Old Bill said:

Go on Ben, you make them!

I wonder if it is really as simple as a pentagonal pyramidal spike inside a square hole? 

(ie. like this https://a360.co/3pagTbb )

And would a hexagon work?

14 hours ago, ruxy said:

A 16 tpi hand chaser will have a handle like wood turning tool/scraper and be at least 2 foot long (and some poor B has to use it) ,  a LH one would not be unobtainium  even going back 70 years +

If you can find them at a decent price, this style of threading tools make a lovely whitworth-form thread in a single operation. 

https://www.ebay.co.uk/itm/284087318863

53 minutes ago, ruxy said:

The draughtsman would probably consult the turner or planner - both IMHO would say avoid a Whitworth form.  It would greatly add to production costs. 

They would cut the thread that they had tooling for. In 1934 that would have been BSW / BSF. They wouldn't have randomly chosen (pre-ISO) metric or UN / Sellers just because it is theoretically cheaper to single-point. 

They might well have chosen to forego the crest rounding for the reasons you mention,  but the thread angle is still rather more likely to be 55 than 60 (or 47.5, or 80). 

54 minutes ago, Barney said:

I feel you are overthinking this, Andy.

Or you are under-thinking it when you suggest that an 1"-16 UN thread is an option. 

I am saying that it is unlikely that the thread form is UN, it is very likely to be Whitworth, and the two are not compatible. 

This has nothing to do with any of the standards, it has to do with the basic shape of the threads, which differ significantly. 

4 hours ago, rog8811 said:

That takes me back, 50 years ago, Guildford tech was the last time and place I used one.  But brings up a the thought of where to get access to a shadowgraph should you want to check out your thread.

I found a YouTube video of someone making their own: https://youtu.be/lw1b58kITxQ

I was wondering if the lead angle of a thread affects the apparent thread angle when projected, but after a bit of 3D modelling, I have decided that the answer is "not enough to cause confusion"

1 hour ago, Barney said:

 1" 16-UN - Unified Extra Fine Thread, which is used for special purposes.
The flank angle is 60°.

But would a 1934 British car be using Unified threads? (Unlikely, Unified threads were invented during WW11) 

I just finished an M77 x 1 thread. I chose that size as I had a 76mm recess that needed a thread, and the 1mm threading tool was the one in the holder. 
If it had been the 1.25 insert then it might have been an M77.25 x 1.25 thread instead. 

41 minutes ago, ruxy said:

the pitch is chosen best suited for the application when you get to dia. above abt. 2"     Things like locking collars for overhead travelling crane trailing wheel axles , multi-auto lathe spindles etc.

Or, indeed for fun. I made an M14.75 x 0.95 thread pair once. Just because that satisfied a mathematical optimum. 

1 hour ago, Doc said:

I agree: measuring the thread form (55 degree or 60 degree) is a good start.

Looking at the thread crests might be informative. Flat crests would indicate that it is _not_ Whitworth form. 

Rounded crests, unfortunately, just mean that it _might_ be Whitworth form at that time period.