Notes on Workshop Techniques




A screwcutting lathe is purpose built to produce accurate screw threads using a single-point tool mounted on the carriage and driven by a series of gearwheels connected to the mandrel. Of course, screwcutting can also be achieved using normal taps and dies, and perhaps less frequently by using form tools (such as die heads and thread chasers).

12.1. Taps and Dies

Hand threading tools.


The advantage of using the lathe with standard taps and dies is that the lathe acts as an accurate workpiece and tool holder, and so the resulting threads are more accurately aligned with the axis of the work than they would be using hand-held tools. However, pitch accuracy and alignment is never going to be as good as with threads cut using a single point tool, so critical work (such as feedscrews) should not be produced by this method. The pitch and depth of the screws are of course a function of the accuracy with which the taps and dies are made. Dies may be of the 'split' type, which offer limited adjustment for setting the outside diameter of the finished thread (not the thread depth), though in practice this may be more trouble than it's worth and there is something to said for using matching taps and solid dies. If adjustment for precise fitting is necessary it would be far better to cut the thread with a single-point tool in the first place. Dies are usually held in some form of holder that can slide along a parallel shaft engaged in the tailstock socket (see here for an example of a tailstock tap/die holder ), whilst taps may be held either in a prupose designed holder or an ordinary tailstock drill chuck. Dies are generally 'sided', such that one side (usually with imprinted information on it) generally has more of a taper to start on the work easily, whilst the opposite side is used to cut as near to a shoulder as possible. It may be found that with a partly worn die a thread can be cut dead to size using the more frequently used tapered side, and when backed off the work the other side of the die takes off another shaving from the thread. If size is critical it is better to release a split die from it's holder (or expand it using the central adjusting screw) before reversing it off the workpiece. In my experience taps and dies from different makers can produce threads which only approximate to their nominal sizes. It takes considerable adjustment of the die before you can be satisfied that threads cut by 'matching' taps and dies are actually a good mating fit. One way out of this predicament is to keep previously adjusted dies in their own turned holder ready to slot onto a carrier rather than have to adjust them each time. Solid dies and taps from the same maker do not suffer from this problem. Taps vary in their construction and employ 3, 4 (and more) flutes either straight cut or helical. The latter form is often used for taps specifically made for machine tapping.

Using a mandrel handle to thread a rod 3/8" BSF with a tailstock mounted die. Work is held in the 4-jaw chuck to avoid it slipping.

I find the pointed end of some designs (i.e.; having a 60 degree male taper on the end) interferes with the ability to thread to the bottom of a blind hole, and so I usually grind the end flat. Don't take this to extremes though, if no lead at all is allowed for you will find that the end teeth will chip off quite easily. Broken taps can be re-sharpened fairly easily on the bench grinder - just be very careful that carbon steel taps are not over-heated in the process otherwise it will lose it's hardness (use a cup of cold water to dip the tap into). Grind a small amount of back relief on the leading 2-3 teeth and taper as necessary to get taper, 2nd or plug tap shapes. Personally, I don't use taper taps at all in the lathe as I find the first few thread tear out, so just use either 2nd or plug.

The high cost of HSS taps and dies compared to carbon types can hardly be justified for the increased performance in everyday use (in my opinion). Having used both in the past I tend to only use carbon taps now for preference. My reasons are that carbon steel cuts just as well (in fact is harder and holds a keener edge for longer) and is easy enough to sharpen when required. The benefits of HSS (an alloy which retains it's hardness at very high temperatures) is hardly relvant to threading. Further, if a carbon tap breaks in a deep hole and it is important to get it out it can either be shattered with a sharp blow, annealed and drilled, or dissolved in acid - all of which are easier with carbon steel compared to HSS.

Return to top

12.2. V Threads

This is the generally accepted method of producing the most accurate threads in the home workshop. A wide range of thread pitches can be produced on the average modeller's lathe, usually covering the range from about 8 to 60 tpi. The threadform of course is dependent upon the tool shape which the operator is free to grind himself, with the most common being V-threads of 55 and 60 degree included angle, and several different types of worm or square threads. One tool for each inc. angle will generally suffice for the V-threads, but worm threads require a specific tip width which usually means making up a special tool for each job.

Small gauges for setting lathe threading tools square to the work, and for checking shape of tool point.


Screw pictch gauges for estimating the TPI of threads.


For V-thread, the shape of the tool depends on the technique used for putting on the cut. I use two methods, the first is to simply feed a tool straight into the work, each sequential cut removing equal amounts from left and right lands of the screw. With a very sharp honed tool, and taking fine cuts (especially in the later stages when the work is near finished dimensions) this method will produce a good finish. Honing the tool is essential in my experience, it doesn't take much work with an Arkansas stone, it just needs a few minutes stoning the cutting edges (use strokes parallel with the tool body - 90 degrees to the movement of the work past the cutting edge). However, there are one or two drawbacks to this method. Firstly, the fact that cutting is taking place on both sides of the tool means that no top rake (or rather, side rake as applied to the cutting edge(s)) can be given the tool, and the top is ground dead flat. This causes crowding of the chips as they are sheared off and flow across the top of the tool, chips are not cleared quickly and may jam in the cut causing tearing of the work surface. This effect is much more noticable with tough materials like drawn phosphor bronze and some stainless steels than compared to free-cutting bms. A good lubricant (soluble oil or straight cutting oil for mild steel, high pressure threading paste for stainless) is essential to minimise this effect. Secondly, burrs tend to be thrown up on the peaks of the threads leaving a ragged edge, and when measured the external diameter is larger than expected leading to an over-sized thread being produced. To get around this usually entails removing the rough peaks of the thread with a fine file or emery cloth (not good if fine accuracy is the objective) - or, using a tailstock mounted die acting as a thread chaser (which defeats the object of using the lathe to some extent). Where I do occaisionally use this method is with a back toolpost-mounted threading tool, and where the thread to be cut is large (in brass or F/C mild steel) demanding the stiffest tool support. I must say though that these days I mostly use a 'set-over' method for screwcutting.

'Set-over' technique for screwcutting.

A set-over technique involves rotating the top-slide to half the included thread angle (say, 30 degrees for a 60 degree thread) thus leaving the right hand cutting edge of the tool (i.e., the trailing edge for normal RH threads) parallel to the right hand land of the thread. In this situation the cut is put on exclusively with the top-slide and the right hand side of the tool does no cutting (see photo). This enables a top rake of about 7 degrees to be given the left hand cutting edge greatly improving the cutting action and hence the finish on the work. This amount of top rake will have no significant effect on thread form. In my case I tend to use HSS for threading tools, carbide tools not taking well to top rake (weakens the tip leading to chipping). It will be noticable in using this method that the chips come away as a clean curl of cut metal and that no burrs are thrown up onto the peaks of the thread. A further modification involves setting the topslide over to just less than half thread angle (say, 25 degrees for 55 degree Whitworth form), such that the trailing edge of the tool just shaves the the right hand land of the thread. This corrects a problem sometimes seen with drawn phosphor bronze rod or stainless steel where scoring can appear on the right hand land. Another solution to the problem is to apply the very last thou or two of cut with the cross-slide (feed straight in) rather than the top-slide. It will be noted by the observant that, when feeding in at an angle with the top-slide, the movement indicated on the micrometer dial does not represent the actual depth of cut. The true depth of cut can either be calculated and read off a table (see here for a set-over table of cutting depths). A rough estimate can be calculated by increasing the cut by 1/8 (for 27.5 degrees) or 1/10 (for 25 degrees).

Even better than relying on setting the cross-slide to a zero setting is to use a cross-slide stop, this being less prone to errors and easier in use. See here for instructions on making a Myford cross-slide stop .

Return to top

12.3. Worm and Square Threads

The main use for this type of thread is where high loading is expected and precise positioning is required, obvious examples being feedscrews and worm wheels. The two main differences from the machist's point of view between these and V-threads are that the pitches are often 'odd' (fractional) and that the actual machining requires a particularly rigid setup. The latter is due to the fact that the effective length of the tool's cutting edge is longer than that for most V-threads, having a wide tip and the thread is usually deep. A worm thread will be designed to mate with a matching worm wheel for maximum power transmission and minimum backlash (or possibly a helical-cut straight gear wheel might be used - depends on the requirement whether for power and speed or just positional accuracy. Example of the former would be a rotary table or reduction gearing, and of the latter a headstock dividing attachment).

The cutting tools for each job will probably need to be specially ground such that the pressure angle (departure in degrees from a 90-degree square thread) and tip width (equivalent to the gap between teeth at their base) are precisely adhered to. The best way of achieving this is to use a tool and cutter grinder such as the Quorn or Kennet, but in their absence a fairly good substitute is a drill/mill with a universal vice and protractor. If you are forced to use a bench grinder you would need to make special tool holders and jigs to grind the correct angles. Probably the best option is to use round HSS blanks (about 3/16" or 1/4" section) to make the cutting tools, a matching holder can then be used and the clearance for the helix angle can be allowed for by rotating the cutter in it's holder. A table of worm thread details is given below*:

                               Tip Width
DP   Linear Pitch    Depth   14.5    20 deg
14      .2244       .154    .0695    .0520
16      .1963       .135    .0608    .0455
18      .1745       .120    .0540    .0405
20      .1571       .108    .0486    .0364
22      .1428       .098    .0442    .0331
24      .1309       .090    .0405    .0304
28      .1122       .077    .0347    .0260
32      .0982       .067    .0304    .0228
36      .0873       .060    .0270    .0202
40      .0785       .054    .0243    .0182
*from: Model Engineer's Workshop Manual; by Geo. H. Thomas

The set-over technique can still be used for cutting worm threads, the top-slide being set to the pressure angle, the cut then being restricted to the leading edge and front tip of the tool. This is a useful improvement over trying to cut the thread using both sides plus the tip, and also means that top rake can be added to the cutting edge to further improve performance.

Owners of lathes with a set of changewheels benefit from the fact that the gear trains required for cutting 'odd' worm pitches can be put together (usually) from the gears already available to them. Owners of machines fitted with only a gearbox have a problem in that these pitches will require the addition of extra gears plus a quadrant to mount them on. The Myford can be equiped for such use either by buying discrete parts (quadant plus gears) or by purchasing the metric conversion kit (or imperial conversion kit if your machine is metric) which will contain all the bits needed.

Myford metric conversion set for the S7.


To cut odd-pitch threads requires that the clasp nut be left engaged throughout the procedure, and if the lathe is not fitted with a clutch this could mean a lot of start/stop/reverse operations which are definitely to be avoided (the motor draws a heavy current when starting so it will likely overheat). The best way out of this is to use a mandrel handle and cut the thread entirely by arm power - neither as difficult nor long-winded as you might think, and an excellent surface finish can be attained this way. If your lathe is fitted with a clutch it is simply necessary to disengage the drive at the end of each traverse, drop the back-gear, withdraw the tool using the cross-slide, and wind the carriage back to the start using the leadscrew handle.

For a worked example of making a matched worm and worm-gear using the hobbing method, see this related article in the 'Projects' section.

Return to top

12.4. Internal Threads

Cutting internal threads offers a new set of problems. Because the tool is working in a hole visibility is reduced, and so is accessibility. In smaller holes (1/4" and less) space is tight and it's difficult to arrange for a sufficiently rigid tool to fit. What was an easy filing job to clean up an external thread becomes difficult with an internal thread, and precise measurement is a problem too - and all this is happening in a swarf-filled hole where you can't see what's going on. If the hole is blind-ended there is the necessity to ensure the tool doesn't run into the end of the bore (I always use a mandrel handle in combination with a carriage deadstop for this job). For these reasons, where a mating internal and external thread are being produced, it is usual to machine the internal thread first and to machine the external thread to fit. Little wonder then that most female threads are cut using a tap! Having said that, with skill, it's possible to machine an internal thread with precision, but it requires reliance on your lathe's micrometer dials and sharp accurately formed tools.

For internal threading of bores over 1/2" diameter it is best to use a round section boring bar with a HSS tool tip held in a cross hole. Click here for a picture of my boring tool holder. The diameter of the boring bar should be as large as possible consistent with the workpiece (around 13/32" for 1/2"). The boring bar holder should allow the overhang of the tool bit to be set to the minimum required to reach the full depth of the thread in the work. I have boring bars down to 3/8", but it's possible to use even smaller bars to about 1/4". However, if my 3/8" bar is too large I tend to use one-piece tools ground from 1/4" square HSS (not the easiest of jobs) or machined from silver steel and hardened. Small commercial HSS jig boring bits (with a 3/8" shank) also make a useful starting point for grinding up threading tools. A quick way of making a threading (or boring) tool from round sliver steel is to chuck a length in the 3- or 4-jaw chuck (use a shim in the 3-jaw between work and one of the jaws) so that it is running eccentric and turn the middle bit round leaving about 1/4" on the end. This will enable the cutting tip to be ground or filed to shape on the eccentric bit. Care needs to tbe taken that sufficient front clearance is allowed for otherwise the tool will rub rather than cut ruining the chances of obtaining a good finish. Best way to check for this is to use a drill gauge (hole plate) and offer up the tool to observe the clearance.

One problem which you might encounter is that of removing a threaded part from the machine only to discover that the thread is undersized. This has happened to me - I relied on the leadscrew collars to give the correct size for an internal thread but did not account sufficiently for flexure of the threading tool. Quite a lot of effort had gone into making this particular part so it was worth the effort of trying to re-mount it in the machine to remove a little more from the threads. In this example the part was in stainless steel 3" long, 1-1/8" diameter for most of it's length but with a 1-1/4" diameter collar on one end. It was bored and threaded 1" x 26tpi, and it was essential that the bore, thread, body and collar were concentric and thus all those operations had been carried out at one setup. How then, to re-set it in the machine aligned so as to pick the thread up?

For external threads its not too difficult to use the top-slide and visually align the the tool with the thread, but for internal threads it's difficult to see the tool tip. In the previous examply I used the following procedure:

That method enabled me to take a few more thou off the thread and yet still maintain the concentricity of the work.

Return to top

(c) Chris Heapy 1996/99.

Back to previous page