The translation of a metric drawing to imperial dimensions presents little difficulty, as the amended dimensions would be in decimal inches, and these would imply the same degree of precision as on the original; thus 26mm could translate as 1", or as 1.02" or depending on the class of work. Translating from inches to mm differs, however, as many designers still use fractional dimension rather than decimals, and some even work in sixty-fourths. If such fractional dimensions are toleranced the degree of precision is indicated. Even so, the fraction itself may imply working in tenths of thous - 13/16" ± 0.001" means 0.8115" to 0.8135", which converts to 20.612mm to 20.663mm. In the same situation a metric designer would probably write 20.50 to
20.5 5mm - a tolerance of ± 0.025mm. The situation is aggravated in drawings of any age, often untoleranced, and even more so on drawings prepared for modelmakers, where the user is expected to relate the dimensions of mating parts and so establish the degree of precision required.
The first step is to go systematically over the whole drawing, converting every dimension to the opposite system. I recommend that figures in inches be taken to 4 places of decimals, or if converting to metric, to three; these can be refined later. Next, seek out and identify the two opposite extremes of precision needed. FIRST, those that must tally with a bought-in component e.g. a ballrace or the bore of a gearwheel which you cannot alter. Mark these plainly and, I suggest, make a note of them. SECOND, mark up figures which are clearly rule dimensions; usually unmachined surfaces, spacing of holding-down bolt-holes and so on.
It is then wise to lay a rule over the castings to assess the amount of machining allowance you have to play with. Note this, in both systems of measurement. You should then identify matching dimensions of the less critical kind where, within the limits of the machining allowance, you can alter the dimension of both mating parts to a round number (e.g. a 15/16" dia spigot (23.81mm) could be probably be made 24mm -or even 25mm - with no problem). The point here, as with the ruler dimensions, is to select conversions that are easy to measure in the opposite system, whether by calipers or with the ••micrometer or vernier.
The next step is more important. That is to establish reference faces and centrelines so that working parts will align. The point is easily appreciated; dimensions in fractional inches will seldom convert to round hundredths of millimeters - the smallest increment that you can work with a normal micrometer. The throw of a crankshaft, for example, or the location of an eccentric relative to the valve-chest, must be dimensioned to an exact match. This must be done carefully and, I suggest, the revised dimensions entered on the drawing as you proceed. Having checked them you can then fill in any remaining dimensions - it is quite certain that some of those previously determined will have to be altered.
My own practice at this stage is to make a new drawing altogether, filling in the metric dimensions (or inch, if I am converting from Metric) as I would were I doing the job from scratch. But there is, in these days of accessible copiers, a very useful aid. Make a copy of the original drawing and go over this deleting all the dimensions with liquid paper (e.g. Tippex). Then get copies made of this undimensioned drawing. You can then enter on one copy the literal conversions you made at the beginning, as well as the reference faces and centrelines. The second copy can then be used for the various steps mentioned, and the third used for entering your final decisions.
I have shown the process on a relatively simple example. This (Fig. 86) is the crankcase, crankshaft and bearing housing for a small two-cylinder single acting high-speed steam engine for a model
boat. I have not shown all the details as by the time the drawing is reduced to book size it would have been smothered in figures. (Most of the omitted detail is concerned with stud positions, oil-holes and the like, which present few problems.) Fig. 86 is the original inch drawing and you will see that I have added reference details; the centrelines of the block and crank, and the length 1/2 inch) of the boss on the right-hand bearing housing, which locates the crank endways. It is, of course, obvious that the spacing of the two crankpins must align with the spacing of the cylinder bores. I have also added machining marks - often absent from drawings published for use by modelmakers, unfortunately.
Before going further there is one decision which had to be made, that of the cylinder bore (and, perhaps, the stroke). A metric design of engine would have a bore of 25mm, not 25.4mm, and if 25mm piston rings were available this is the proper conversion. In this case, however, the piston rings were to hand for a bore of 1.000 inch; to reduce the bore to 25mm would mean filing at the gap, removing 1,26mm (about 0.050") and though this presents no difficulty there would be a risk that the ring might not fit properly. So, the decision is made to keep to 1" or 25.4mm and as there is no difficulty with the crank either, the throw is retained at the metric equivalent of I/2 inch - 12.7mm.
Fig. 87 shows a second copy of the drawing dimensioned with exact metric conversions, though all have been rounded off to the third decimal place. You will notice a number of dimensions marked with an asterisk *; these are all rule dimensions, which can be altered within the limits of the machining allowance (3/32 inch or 2.4mm)
to suit the machinist's convenience. We have one dimension (apart from the cylinder bore) which must fit a bought-in part; the 7/16 in. dia on the shaft, marked A, carries a commercial bevel gear. Compare Fig. 86 with Fig. 87, and you will see that it is dimensioned at 11.11mm (0.437in.). The machinist can work to either, the difference between the two being 0.003mm or 0.00001", not significant. There are some places where parts can be machined to a mutual fit, marked B. The end bearing housings, both the spigot and the bearing bore; the mating part of the shaft; the ends of the shaft, which fit to flywheel and coupling; and the vertical reamed hole which carries the bevel-wheel shaft. Consider these in turn. The spigot can be 42mm, not critical provided the crank can be passed through. The shaft bearing can be 10mm, a standard reamer size; a little homework shows that the combined bending and torsion stress will still be reasonable. The shaft ends can be 9mm with no harm, still standard reamer size, and the vertical shaft 8mm. Now consider the dimension C - the height of the block above the crank centreline. This dimension would affect the piston clearance at dead centre. Examination of the connecting rod forging showed that centres had to be held fairly close. The exact 73.025 is, therefore, held at 73mm, instead of being altered to 75 which would be a more normal dimension here.
Now for the critical part - aligning crank and block. Compare Figs 86 and 87 during the next stages. The cylinder centre distance, D can be refined to 38mm, and that from the centre of bore No. 1 to the reference face made 26mm without overrunning the machining allowance - D,. (We
Was this article helpful?