I may be getting the hang of this milling thing. Today I got most of the way through making the crank for the beam engine. This was two setups and more tool changes than it probably should have been but I am pleased with the result. I just need to mill off the back of the part to separate it from the block of stock I used as a handle.
It looks like it should even if it may not exactly match the drawings. I realized that the only things that really matter are that the two holes are parallel, the correct size and the correct distance apart. Everything else is mostly shaping things to resemble the cast part a full size engine would have. As long at things are symmetrical, nobody will notice if the end radii are a bit too large or small.
I also improved my finding of the center of the rotary table by using a new gadget, a coaxial centering dial indicator. This device gets chucked in the mill spindle and you center on the hole while the mill is spinning (at low RPM). This is actually fun as opposed to the usual dial indicator spinning holder that makes you keep having to move around to see the readings.
The other improvement which is much less photogenic was the acquisition of some drill blanks. I used a 3/32″ dill blank chucked in the mill to center up the part on the table center for the second setup so I could round the small end. I previously just used a reversed drill bit and I don’t think it produced as accurate results. The drill blank is both more rigid and has not flutes to mess with the alignment.
The title seems a little silly for this one since most everybody has at least seen calipers if not used them. They are used for a reasonably precise method (I have already covered at least one unreasonably precise method) of measuring the distance between two surfaces in a quick and convenient way. Presuming you can find them, of course. Note that no vernier scale caliper is pictured above even though I own one…
The most used measurement is the outside one. I will often use calipers for marking out a part as well as measuring an existing object to find out what size it is. Handy if you have unlabeled metal stock lying about. In the picture, we can see that the shank of my 3/8″ end mill is about .001″ under sized according to my inexpensive caliper. If I needed more precision, I would get out a micrometer.
Using the other set of blades, one can measure inside distances between flat surfaces. It turns out my 21mm wrench is actually a 21.11mm wrench.
Calipers can also be used to measure depth in two different ways. The first is what I have used for years, the other end of the caliper.
The second is one I only learned about recently and while it is somewhat specialized, it allows a more reliable reading of a step using the end faces of blade end. I held it above the piece for what I fondly hope is clarity.
The alternatives to calipers are fairly well known and are all a tradeoff between precision and convenience. There are probably people out there doing fine work with just a rule or just micrometers but they are giving something up.
Calipers come in Vernier, dial and digital models. Vernier versions are the most robust and proof against fluids (cutting coolant) but slower to read. Many have both Imperial and metric scales on them. Dials are quick and easy to read but only come in one set of units. Digitals can convert between Imperial and metric, can be zeroed at any point and can give implausibly precise readings. And they need batteries which is why my pictured one is not getting used for any of the action shots.
Calipers come in a vast range of quality from very cheap plastic models (1.99!) to very uncheap ones you would be careful not to breath on. They also come in lengths from 6″ to 24″, at least at my local tool supplier which lists 83 different models ranging in price from $50CDN to over $500. Mine are much more in the $50 range (probably less) but do a good enough job anyway. I use them to get close and then check final dimensions with a micrometer if it matters.
This is one of those tools that you don’t need until you do. As evidenced previously, I have managed to turn out some parts that are less perfect that one would hoped. While the list of possible sources of error is long, a fundamental part of setting up machine tools is levelling them. Even a solid casting like a lathe bed or a milling machine base can acquire detectable twist in it when you bolt it down. To take out the twist, you put shims under the low corners or some other adjustment until the measured surface is level. To detect twist, you use a precision level.
I have avoided going through what is almost certainly a fiddly process with either of my machine tools but I decided the time has some so I acquired the Starrett level depicted. The certificate in the box asserts that each division on the tube represents .005″ difference in elevation over 0ne foot. This ought to do it.
Initial readings on the mill Y-axis ways (which are the ones that are part of the base casting suggested that my previous attempts with carpenter’s level were less than perfect. At least I didn’t buy this thing for nothing.
Fortunately, the stand has leveling feet so off I went with a wrench and a lot of bobbing up and down and cranking. It turns out that one of those .005″ marks represents about a half turn on a foot. Presuming that the other foot is on the floor… I need to develop a method that avoids trying to level an inadvertent tripod.
After a lot of faffing about, I got to about one mark of level and at least the back and front agreeing on which way that out of level goes. I called it done for now because I was chasing that mark around the corners. I need to do some more reading up but I suspect that this is where the shims come in. The top of the milling vise show just about the same reading so nothing inherently funny is going on.
There are methods of detecting twist in a lathe that don’t involve a level. Twist in a lathe bed manifests as a taper in what should be a parallel turned surface. You can turn two ends of a round bar and measure the result with a micrometer. You then shim whichever tailstock corner is low. Repeat until both ends come out matched.
You can also buy a pre-turned bar, mount it between centers and use a dial test indicator to compare ends. At least I think that’s how it works.
Presumably, you can carry out the same sort of process with a mill but given the way the mill table takes up most of the space, it would be a pain to say the least. You can’t trust the spindle or column at this stage so I think you would have to remove the table and sweep the ways with an indicator mounted on the ways. Suddenly, a fancy level doesn’t seem such a bad idea!
This is one of those purchases that will get seldom used which is why I avoided it as long as I did. My “budget” model Starrett level was about $180CDN. You pay for greater length so my 6″ model is the shortest I could get away with. The 18″ one is about $1200.
There are digital machinist levels that cost more than the analog ones. I don’t know if they work as well. They claim similar resolution and I even found one online that has an app that relays the readings. That would be handy if you wanted to avoid jumping up and down to check as you adjusted things. It would be a huge time saver if you were levelling a CNC machining center, for instance.
At first glance, a vise is not a measuring tool so why am I talking about this thing? Well, a toolmakers vise is a vise (duh) that is precision ground on just about all of its surfaces, certainly all six square sides, the jaws and ways are square and parallel to a high degree. The depicted vise is also a “screwless” design that uses an hex draw bolt instead of the usual screw to secure the moving jaw in both an inward and downward direction. This avoids introducing error due to jaw lift.
Suppose you have a part that you want to check for parallelism of some awkward surfaces. As an example, I have just made some t-nuts for securing things to my milling table. I was wondering how I did with my setup. An actual metrology application or, as a certain Youtuber says, “the surface plate: where dreams go to die”. As you can see, checking the vertical surfaces of the inverted T is not just a matter of plunking it on the surface plate. The darned thing won’t stand up on that face.
And so the vise comes into play. I know that it is very square and so if I clamp the t-nut in the vise, the clamped surfaces should also be parallel to the corresponding vise surfaces. If I check both ends, it will tell me how off it is. To take the reading, I leave the indicator and stand stationary and slide the vise and part around under the indicator tip.
And the answer is, in machinist terms, quite a bit, about .0015 if I am generous. The same error is present on the other nut I made so at least I have repeatably created the error. This is why one makes t-nuts for practice. The required precision is not great and you can mess things up and still have a usable result. If it was important, I would have to investigate the source of the error (milling vise alignment, mill head nod, etc…)
As is probably obvious, you can also use a toolmakers vise for workholding. I have seen various videos of this sort of vise being used to hold a part being milled.
Toolmaker’s vises come in different designs and many sizes. Mine is only about 2.5″ long overall. They can get expensive really fast.
As an alternative, you could clamp the part against a 1-2-3 block and rest it on the plate. Or use a v-block or some other way to prop it up vertically. In this case, you would likely need to move the indicator stand around which is finicky in that you can tilt things and mess up your alignment.
As always, you can spend a lot of money on a toolmakers vise if you want to. There are some beautiful makes and models out there that are breathtakingly expensive and certainly more than most hobbyist need. The cheapest 25mm/1″ version I found online goes for about $60CDN. Mine is a somewhat higher quality model but still accessible at least in this small size.
Telescoping gauges are for measuring the diameter of holes, the width of slots, and any other gap that is impossible to get calipers into. The standard model is T-shaped with the bar of the T composed of two sprung arms with round ends. The other end of the handle has some sort of locking mechanism. You stick the appropriate sized gauge into the hole, lock the arms, pull it out and then measure the result with a micrometer or calipers.
I say appropriate size because each gauge has limited range of measurement and thus telescoping gauges come in sets or at least collections. Mine is a motley assortment bought from a used tool store.
For small holes, there is a different style of gauge that is a split cylinder with a center plug that adjusts the split. I only have one of those.
There are digital and dial direct reading bore gauges which are the more precise but more expensive choice.
The basic import sets go for about $40CDN and cover a range like 5/16″-6″. You can spend more for better ones just like other measuring instruments but for the money a dial bore gauge set would be a better buy, I think. I don’t have one of those because I have not had cause to use my telescoping gauges enough to start wanting better.
I consider this sine bar to be the most exotic measurement tool I have. I bought it mostly for fun. A sine bar consists of two cylinders of the same diameter fixed with their centers a known distance apart. The top of the connecting bar is parallel to the line between the two centers. All this done with as much precision as you are willing to pay for.
What the bar is used for is to “construct” precise angles between the bar surface and the surface the bar is resting on, typically a surface plate. Given the desired angle, one does the appropriate calculations (hence the sine name) to get vertical displacement for one end and builds that height out of gauge blocks.
Here as an arbitrary example, is a setup to get an angle of 32 degrees, 16 minutes and 27 seconds. Reference to an online calculator produced a displacement of 2.670″.
I called this tool exotic because it is difficult to conceive of a circumstance where I would actually need to measure to set up an angle this precise. I can use it as part of a machining setup or to check something like my engineer’s protractor for accuracy. I expect that an experienced machinist has more uses for one so who knows.
Budget sine bars cost less than $50CDN new so it wouldn’t break the bank if you decide you needed one. There are lots of protractor variants that are probably more practical for most jobs.
The “sine tool” family includes plates and vises. The plate is just a wide bar. The vise is a vise but includes some sort of angle locking mechanism and the two cylinders to allow precise angle setting. Like a lot of precision tools, spend as much money as you want for increasing precision well in excess of the average hobbyist’s needs.
Just like surface plates are a standard for flatness, gauge blocks are the machinist’s standard for length. And also like surface plates, they are less simple than they appear.
Created from well conditioned steel and polished to a very find degree, gauge blocks come in sets of different sizes that are combined to get whatever length is required. An imperial set is typically 81 blocks and covers ranges from .001-12.000″ in .001 increments and .2000-12.0000 in .0001 increments. The odd starting point for the ten thousands is because the first block is .1001 or such since nobody can make a durable block .0001 thick. Making up the required combination may count as playing with blocks. 🙂
The assembled length is used to compare or calibrate whatever you are trying to evaluate against the standard. My singular “real” usage so far has been checking the accuracy of various used instruments I have purchased. As I mentioned somewhere previously, one can mount a dial test indicator on a height gauge, zero it out on the gauge block(s) and use it to check parts for expected length.
A fun feature of gauge blocks is that you can “wring” them together. The highly polished surfaces allow a close enough fit that they will stick to each other. Very handy when you need to move a pile of four blocks.
If you don’t have a gauge block set, you can start with a micrometer and measure some object, take the result and compare it to the resulting measurement from other instruments to see if they agree. Some higher end micrometers come with a block usually called a “standard” which is used to check the mic. This is effectively a single gauge block.
Cheap gauge block sets go for about $180CDN new and things go up from there. In my case, I opted for a used set of some personal and historical interest which I will now enthuse about:
In the 1800’s, New England was a center of industrial excellence with the precision manufacturing aspect perhaps most famously represented by the Waltham Watch Co. Where you have machinists, the tool manufacturers follow and so it was with Massachusetts. That industrial sector is a shadow of its glory days but the shadow is a long one with the L.S. Starrett Company still in business. I have more than one Starrett measuring tool but my gauge block set is a product of a less well know competitor, Van Keuren.
The Van Keuren Company still exists as a brand but no longer operates out of its building in Waltham, MA. (As an aside, I worked in the Waltham area for about six years, love those brick mills) I know they still were there in 1942 because that is when my block set was manufactured. The three digit serial number suggests either an early product or one that didn’t sell a lot.
The set sill contains the inspection certificate with the deviation from labelled size in millionths! for each block. I do wonder if they still conform to those results but am unlikely to need to care. 🙂
Next to calipers, the most common measuring tools used by machinists are dial indicators and dial test indicators. These precision instruments are used to precisely measure changes of dimension over relatively small ranges. They consist of a graduated dial with a movable face and a “hand” driven by some sort of probe that one applies to a surface. They are used while held in some sort of holder rather than in one’s hand. Here is a typical holder setup:
I am lumping the two indicator types together since it makes it easier to explain how they differ. The major distinction is one of range. The big one in the photo is a dial indictor: it has a plunger probe with an overall travel range of 1″ with each revolution of the hand covering .1″ or 100 thousandths which are the smallest units on the dial. I use my indicator for “rough” measurement such as getting a milling vise aligned or centering stock in a chuck.
The other is a dial test indicator. It has an overall range of only .030 inches and is graduated in .0001″ increments. Test indicator dials also tend to be marked counting up from zero in both directions and, like this one, have a lever probe instead of a plunger. Test indicators are typically used for really precise measurement such as checking the parallelism of two surfaces. Here is a setup I used to see how parallel my 1-2-3 block was :
I moved the block around under the probe and got at most about half a tenth movement. I may be detecting fingerprints on the polished surface. I am calling it good enough. 🙂
Indicators do come in digital versions but the vast majority are still analog dials. They come in both metric and imperial versions although in North America the vast majority available seem to still be imperial.
Indicators also come in a wide array of price levels. My dial indicator is a cheap one that cost maybe $50CDN with the stand. My dial indicator was purchased used and is not a cheap one, being made in Switzerland. You can easily spend several hundred dollars on an indicator but if you don’t have a pressing need it is probably not worth it. (I just liked the idea of measuring things in .0001″ increments. I have no other excuse)
There aren’t really genuine alternatives to indicators but cheap ones are readily available that seem to be quite good enough for most work. If you aren’t doing machining, you probably don’t need on unless you want to check the runout on your drill press or assess the flatness of machined objects.
This week I am talking about the humble block of stone known as a surface plate.
What Is A Surface Plate For?
When evaluating the dimensional properties of various machining related items, one quickly arrives at the need for something to compare to. A standard or reference as it were. One of the basic results one hopes machine tools produce is straightness and it’s two dimensional version, flatness. The surface plate is the general standard upon which evaluations of flatness rest. For what is basically a simple block of rock, there are some details I find fascinating.
As I mentioned last week, you can set things on a surface plate and evaluate how parallel a top surface is with the side resting on the plate. With additional tools, you can also check for squareness.
There is also a way to check for flatness of a surface. One applies a marking fluid (ink/paint) to the plate and rubs the surface to be evaluated around on it. The result is paint on the high spots. What you do with that is whole other post but basically you can scrape, grind or machine things to improve flatness. Scraping is the basic manual way of getting something really flat and a skill I hope to acquire eventually.
Modern surface plates are commonly made of granite although cast iron and glass ones are also made. (I have never seen either for sale by vendors I frequent). Granite plates are produced and serviced! by a variety of vendors big and small. The history of surface plates is an inversion of the usual progression of technology because they were originally made from cast iron and granite took over when wartime metal shortages made trying alternatives such as glass and granite attractive. It turned out that granite was good enough for most uses. Cast iron is apparently preferred for really high end work.
Plates come in an assortment of sizes from 6″x 12″-ish up to huge multi-ton monsters. As you would expect, with greater size comes greater weight and things quickly get out of reasonable reach of the hobbyist. My plate is 12×18 and 3 inches thick and weighs 72lbs. I would have gotten a smaller one but this was on sale.
Speaking of on sale, plates come in a variety of grades with increasing levels of flatness. Mine is alleged to be flat within .0001″ which is quite good for an inexpensive (~$60CDN) but I have my doubts. Plates come with inspection certificates but the vendor didn’t bother to actually fill it out… I am not too worried because it is flat enough that I can’t detect any variation using the best methods I have available, a dial test indicator. (Teaser for next week!)
Unless you are aspiring to finicky levels of accuracy, a surface plate could be overkill. A 70lb block of granite is not something you just whip out of a drawer for a quick measurement. The best alternative I have used in the past is a piece of plate glass, in my case a discarded glass shelf. It still sits handy to my workbench for use as a reliably flat assembly surface.
I have also seen at least one Youtuber using a piece of granite counter top. I have no idea how flat that is relative to an official plate but with modern processes, probably close enough for many. Actual granite surface plate manufacture involves leaving the blanks in temperature controlled rooms for months for the internal temperature to even out to avoid the minute distortions it causes. I doubt countertop companies do that except by accident.
Formally, metrology is the scientific study of measurement. Machinists typically use the term to refer to actual measuring. This is an aspect of metal working that I find fascinating, probably because there are all sorts of wonderful tools involved. To predict that I will be overkilling this particular aspect of the hobby takes no special foresight. 🙂 I have resolved to share this enthusiasm through a series of posts.
There are various aspects to the subject I intend to explore besides the tools themselves. I briefly touched on engineering fit in my roller gauge design. That is a way of establishing one required measurement. There are also techniques I need to learn more about; I see machinists on Youtube taking measurements in clearly deliberate motions I don’t yet understand the point of.
Depicted in the photo are my new surface plate and height gauge. Next two series topics sorted!