I offer this for those who like to work with lathes, and who make an effort to produce accurate work. A couple of years ago I made a small quantity of arrow heads, target heads, not hunting broadheads, for a guy in the UK. He was a BA crew member. They stay at the Lord Charles Hotel in Somerset West on their Cape Town stopovers. The guy rocked into Allied Arms on the off chance that we would have them or could make them. Apparently they are expensive to buy and expensive to have made in the UK.
The following is the text of the letter we enclosed with the heads when we posted them. As you will see, I found it very difficult to achieve exact repetition, but I think I did quite well. The big lesson, however, was that such repetition within close limits is not easy, and I now understand why machine shops in the UK charge so much for this kind of work.
I never got a reply to my question as to how the heads performed, which I thought a little unkind considering the amount of work that went into them for a peanut price.
Dear Mr —-,
We enclose herewith your arrow piles, with our further apologies for the delay.
Whilst the difficulties we encountered are not of direct interest to you, we believe certain aspects are worth mentioning in the interest of completeness.
Quality
Although we had only one sample, we got the impression that these piles are made to close tolerances on CNC machinery. The sort of consistency obtainable on such equipment is rather more difficult to achieve on a manually operated lathe. Consequently, we cannot claim to have achieved precise dimensional consistency from one pile to the next, and you will notice in particular slight differences in the position of the annular groove just behind the conical point. That’s because there are slight differences in the length of the piles, and the position of the groove is measured from the base not from the point.
Nonetheless, we believe we achieved sufficient consistency that the differences will be too small to show in their performance.
All were measured and weighed. The shortest and longest are 30.90mm and 31.00mm, which is a variance of 0.10mm or four thousandths of an inch. The smallest and largest diameters are 7.015mm and 7.025mm, which is a variance of 0.01mm or four tenths of one thousandth of an inch. Minimum and maximum weight are 67.70 grains and 69.00 grains
The sample measures 30.70mm long, 6.99mm diameter, and weighs 68.80 grains. Close inspection showed that the nose had been slightly flattened, presumably from impact on something harder than a target. We concluded therefore that the true length was slightly greater, hence our estimate of 31.00mm. In the event, the shanks turned out fractionally longer than the sample.
The diameters were machined 7.02mm simply because we made our collet tool slightly oversize, and the shank of the pile must be a close fit in the collet to ensure concentricity of the conical nose and the tapered tail. 7.02 is only one thousandth of an inch bigger than 6.99mm and we considered that close enough not to necessitate remaking the collet.
The weight of the sample is within the 1.30 grain variance of the thirteen piles produced.
Just for the sake of interest, there are thirteen because fifteen were made in case some were lost because of production errors, and in fact two were.
Finally, there is slight variance in the surface finish, but in general it is equal to the sample.
We are confident that it would be possible, for the manufacture of larger quantities, to halve these tolerances or even better.
Material
At first we thought the sample was aluminium or aluminium alloy, because that’s what it looked like on visual inspection, and because it could not be attracted by a magnet. Being such a small and light item, it was not heavy enough to discern what it was made of just by feel.
Consequently the first copy was made of aluminium. We were surprised to say the least when it weighed only 40% of the weight of the sample. The weight difference indicated that the sample was steel, which seems to be borne out by the weight of the finished piles. We are somewhat perplexed therefore, that while our steel piles are attracted by a magnet, the sample is not. So what is it made of? It can’t contain any aluminium, or it would not weigh the same as our steel piles.
On a slightly different angle, we think silver steel might be a better material from the manufacturing point of view. We made the piles from free cutting steel which is the easiest steel to machine. But we found that the most time consuming operation was turning to the exact diameter required. Silver steel is precision ground which would obviate that operation.
It is however, otherwise difficult to machine to a good finish, and also has an affinity for rust so it would have to be blued or plated. At any rate it would need further investigation.
Rust Prevention
As noted we made the piles from steel because that’s what we thought the sample was made from and because the finished piles weighed the same as the sample. The fact is however that unprotected steel rusts especially in an outdoor environment. We have coated them with light oil to prevent them rusting in transit, but you will have to ensure rust prevention thereafter. We suggest cleaning them and oiling them for storage after each use. That is regarded as standard practice by those of us who use firearms.
The holes which accept the shafts should be cleaned of oil before the shafts are fitted. In particular, if the shafts are glued in, the oil will prevent adhesion if it is not removed. Dip the folded end of one of the pipe cleaners enclosed with the piles in paint thinner and swab out the holes with it. Use the other pipe cleaner to dry them out, not that they won’t dry anyway, as thinner evaporates quickly.
We could have blued them, which would have provided some protection. However, it would probably wear off quickly from the friction of target penetration and withdrawal. A better method would be nickel plating which is very durable, but it could raise problems with the fit of the shaft into the pile, as it would be necessary to match plating thickness to drill diameter, which would require a consistency of plating thickness that would be difficult to maintain.
With the advantage of hindsight, we would suggest that these piles would be better made of brass. It’s much easier to machine and of course is totally rust resistant. The machined finish of brass is matt rather than bright, but that shouldn’t be important.
It’s also about 15% heavier than steel, but the same weight could be achieved by drilling a smaller diameter hole forward of the shaft hole, towards the point. That would be possible because the shaft hole is quite short, leaving a lot of solid material in the nose.
Conclusion
It has been an interesting exercise. Considerably more difficult than expected, let it be said. But we have learned much, one of the main lessons being that precision machining of such things as gun parts and jigs and tools on a one off basis is very different from producing a number of identical items to the same precise dimensions.
It’s not that one is more difficult than the other. It’s the relative importance of time. One can afford to be lenient about time when making a single component or a tool that may see repeated use. But series manufacture must be rapid to be economic, and we found it impossible to achieve the necessary production speed.
Consequently we now understand why the machine shops in the UK charge so much for such work. The reality is that only substantial quantities produced with special tooling are viable.
We would not be averse to producing more of these but only at a higher price and at least 100 pieces per batch. Otherwise it’s not worth setting up the lathe to make them.
Finally, we would very much like to know how our piles perform, whether good, bad, or indifferently.
Yours faithfully,
Dick Boothroyd.
[Originally posted to SATalkGuns -- Admin]