Think & Tinker, Ltd.
CNC router bits and carbide cutting tools with tutorials and technical guides

Selecting the Best Grade of Collet for Your Application

Unfortunately there is no universally accepted format or protocol for specifying the quality grade of a collet, or of any toolholder for that matter. ISO standards not withstanding, the actual names applied to different precision grades seem to be so randomly chosen that they offer no help at all in comparing products from one vendor to that of another.

Such is the also case with the grading system that we have adopted. However, in a effort to lend some credibility to our naming conventions, we have established performance based criteria that can be used to decide which level of precision most nearly meets your minimum requirements. In the tutorial "Measuring Spindle Runout", we discussed the 2 major types of static runout (TIR) and how they could be measured. What was missing from that discussion, as well as from the published specifications of most tooling vendors, is where, relative to the face of a collet, the TIR is measured.

Special Note on Collet Reducers
  • ANYTIME you use a collet reducer to mount bits with different shank diameters in your spindle or router, ALWAYS assume that the TIR of the reducer will add to the TIR of the collet in the worst possible way.
  • Base your decision on which grade of collets and reducesr to buy on on how much TOTAL TIR your machining process can tolerate.
  • A great reducer will NOT make a crappy collet better!
  • A great collet will NOT make a crappy reducer better!

Generally speaking, collet manufactures specify a single value for the runout (TIR) of a collet. Often it is not clear where, relative to the collet face, the measurement was made. International standards for ER style spring collets (DIN6499 and ISO15488) specify that the TIR of a collet should be determined at the collet face and at a distance not less than 4 times the clamping diameter (D). For example, the TIR of a collet with a clamping diameter of 1/8" (3.18mm) should be measured as close to the face as possible and at a distance not less than 4 X 1/8" = 1/2" (12.7mm) along a certified precision calibration blank.

The measurement reported is generally the one taken at the 4 X D point on the blank. No mention is made of the other measurement, or the phase angle between them. Without these 2 pieces of infomation, there is no way to determine what the TIR will be farther from the collet face. It may not souund like much of a problem but, if the two measurements are exactly 180° out of phase, a tool clamped in the collet will be skewed across the central axis with the TIR increasing the farther you get from the crossing point. For collets used with microtools (like ours), it is more important to know how the tool is "wobbling" at the length where it is actually cutting. For 1/8" shank bits with an overall length of 1-1/2", this is typically about 1" from the collet face.

For that reason, we denote our collet grades with a TIR specified at the collet face (TIRF) AND at a point 1" down the calibration blank (TIR1") as shown in the image on the left. These values are listed in the table on the right.

Grade TIRF TIR1"

All other things being equal, assuming that you are using a well made collet with consistent clamping pressure, proper balance and reliable bit retention, the Total Indicated Runout (TIR) is the collet parameter that has the most direct effect on the quality of your cut and how long your bit survives. In extreme cases, it can also affect the life of the bearings in you spindle / router, but that is a situation deserving it's own discussion.

When you are evaluating a collet for addition to your tool kit, a number of factors should be considered before selecting the collet grade (and the amount of money you will need to spend). The most important ones are:

  • When solid carbide tools, with a cutting diameter less than 1/8" (3.18mm), first began to appear, users were justifiably obsessed with runout. The early materials were relatively brittle and not very tolerant of excesive transient side loads, the kind of high-frequency sinusoidal loads that can result from TIR. Even though the early carbide tools could demonstrate a surprising amount of flexure, their relatively "delicate" nature compared to high-speed steel (HSS) led to the wide spread belief that they should only be used with very precise, low TIR spindles. As carbide development progressed and tougher compositions began to dominate the market, the reputation for delicacy nonetheless persisted. The advent of user-friendly CNC machining systems has, in some measure, been pivotal for changing this perception. Today's carbides are tough. In some cases, almost as tough as HSS. However, you hould still do everything you can to minimize the runout of your system when using small diameter tools. A good rule of thumb for end-mills, and other side-cutting tools, with diameters less than 1/8" (3.18mm) is that TIR greater than 10% of the cutter diameter can be fatal to the tool. Initially the flexibility of a tool, cutting with high TIR, will keep the tool alive, but as soon as the edges start to dull, PING. Put another dollar in. Even if the tool does not break, a lot of runout can make cuts much larger than intended and spoil the beauty of your project.
  • An exception to the above rule occurs when you are using a very large tool whose cutting diameter is more than 2 times the maximum shank size your router/spindle can accommodate. The rotational mass of such a tool (e.g. millwork router bits and large diamete "V" tip carving tools) is so large that excessive TIR will cause such violent shaking, that the bit can work it's way out of the collet and go flying off into the wall (hopefully not into your face). For applications where these types of tools are commonly used (e.g. router tables), use UP or EP grade collets. You will be more than a little surpised at how much noise reduction you can get from a properly designed and manufactured collet, not to mention the degree of improvement in the edge finish.
  • By now most people have heard the expression "putting lipstick on a pig". The best collet in the world will not make an old router, with worn out bearings, magically better. A good collet is also wasted on a router if the TIR of the tapered bore is already too high. That is why we always recommend that you measure the TIR of your spindle before you sink money into a precision collet set. Generally speaking, if the TIR of the tapered bore is <0.0002" EVERYWHERE along the taper surface, your router will probably benefit from one of our collets.
  • The thing to remember is that TIR is additive. Any collet TIR will be added to the spindle TIR, unless by some miracle you manage to rotate the collet inside the bore just the right amount so that the TIR contributions subtract from each other. Since this is unlikely to happen on a regular basis, you can be sure that, sooner or later, TIR cutting artifacts will show up in the worst possible place in your design.
  • Pine, fir, cedar and other softwoods: If you are machining softwood, or any other material that tends to compress during the cut and then spring back after the bit has gone by, using a collet grade higher than PG doesn't really make much sense. There will probably NOT be a noticeable improvement in the quality of the cut using a lower TIR collet. For most projects with small tools, SG collets will probably be perfectly adequate.
  • Hardwood and plastic: Although most hardwoods and plastics do not display as much compression and spring-back as their softer kindred, a PG collet will almost certainly be more than adequate for most projects. As above there will probably NOT be a noticeable improvement in the quality of the cut using a higher grade collet.
  • Copperclad and Metal-laminated substrates: Whenever you are cutting a laminated structure with outer layers of metal, an EP, or at least a UP grade collet should be used. Most of these laminates feature outer layers of fully annealed metal which will produce significant burrs in the presence of excessive TIR. Specialty cutting fluids are available which significantly reduces the generation of airborn particulates and minimize the formation of burrs in the metal layers.
  • Shell, marble and soapstone: Cutting natural shell products, softer minerals and any other materials that are brittle and tend to spald, or chip, requires that your system have very little runout. However, as long as the TIR is low enough that the bit is not in danger of decapitation (TIR < 3% cutter dia.), bit flexure will absorb enough of the wobble that it probably wont be noticible (except to pool cue collectors with eye-loops). For these materials, PG is normally adequate, UP if cutting small, precise details in natural shell.
  • Aluminum, brass and copper: Once you start cutting metals, even the relatively soft non-ferrous flavors, you should seriously consider upgrading to UP collets. Especially if dimensional fidelity and the quality of the edge finish are going to be important to the final product. If at all possible, use a high-quality, water-based cutting fluid to keep these metals from welding to the cutting flutes and breaking your bit (not to mention trashing your surface finish). Don't depend on "rules of thumb" for selecting a good speed (RPM) and feed rate. Invest in a good reference source (like G-Wizard Calculator from CNCCookbook) or join CNC Zone to get help from some of the smartest people on the planet.
  • Mild steels andstainless steel: In my opinion,when you are machining steels a cutting fluid is not really optional. For these materials, UP or EP grades are best. Again, a good speeds & feeds reference is highly recommended. Better yet, talk to a local machinist or become a member of CNC Zone.