<< PREVIOUS NEXT >>
Metal Scraping: Preview
1. METAL SCRAPING. HOW TO DO IT, AND HOW TO BECOME PRETTY GOOD AT IT.
In preparing this material, my goal has been to make it useful to both the beginner and the experienced worker. . .
2. WHAT IS IT LIKE TO SCRAPE. Spread on your surface plate a thin layer of High Spot Blue or other marking medium, lay on it the workpiece to be scraped, and slide it a few times, with moderate and even pressure, in a circular or figure eight pattern one or two inches across. Remove the workpiece. Its higher areas rubbed on the surface plate, became marked in blue, and are clearly visible. Now . . . with your scraper shave away the metal at those marked spots. The surface . . . will have become a little bit flatter, and you will have just made your first scraping pass.
Repeat this cycle as many times as needed . . . After a sufficient number of passes, the surface will have become as flat and as smooth as you want to make it. That's all there is to it.
Well, maybe not quite, especially at first . . . but that is just about the way scraping will become, once you have developed a moderate knack for it. So let us see how to develop that knack. First of all let us look at what you will need to get started and have an early hands-on experience. Then we will consider the various aspects of scraping . . . and everything you will need to be able to do quality work. I will try to explain not only what to do, but why you are doing it, and how to recognize early if you begin to veer off course.
We will begin by using steel scrapers,
which is the simplest and least expensive way . . .
Carbide scrapers are sharpened less frequently and in a few moments, so they save
considerable time . . . They can be shop-made inexpensively, or purchased ready made. However to get
set up to use them properly, some specialized equipment is needed, which can be expensive.
Forthcoming material, presented in the previous page, will
describe how you can grind, lap, and also make your own carbide scrapers, and how
you can build in your shop, affordably, the equipment needed to do so: a Diamond
Wheel Grinder, a Lapping Machine, and a range of accessories to use with them.
3. GETTING READY. Here are the things you will need. They will be sufficient to do a wide range of scraping, including most of the examples described at the end. Let us consider each one in some detail.
SURFACE PLATE. It is a compact block of granite that originally came out of a volcano, then became part of a mountain and sat there for several million years, waiting to be collected for you. So you can trust it to be dimensionally stable. It contains a considerable amount of quartz, so it is very hard and durable, and the working surface has been made extremely flat. . . I recommend you get a 9"×12" "Toolroom Grade B" . . . which, although it is the lowest grade, is flat to a remarkable .0001" overall . . .
SURFACE PLATE COVER . . .
SCRAPERS . . . are available inexpensively . . . are well made, and require only a moderate amount of grinding to be ready for use . . .
SHARPENING STONE . . .
LAYOUT PAD. There are many ways to spread the marking medium, usually High-Spot Blue, on the surface plate. I find a shop made fabric pad most convenient to spread it in a controlled, uniform, and reliable way, while keeping my hands reasonably clean. It is superior to rollers when a thin layer of blue is needed.
PRACTICE MATERIAL . . .
GRINDER. I am assuming you already have a bench grinder to grind your HSS lathe bits. It will also work well for steel scrapers, if the wheel is the "cool", white, friable kind . . .
NICE TO HAVE. The next items are not essential, especially at the beginning. But they are convenient, and if any of them happens to come your way now and winks at you, and your budget allows it, you may want to grab it . . .
BEVEL EDGE SQUARE . . .
MAGNIFIER. It is very useful for looking at the cutting edge of
your scrapers, especially while you are getting the hang of sharpening
them, or whenever you want . . . In front of the photo
is an 8× comparator. The color corrected
(achromatic) kind is more expensive but gives a better image, especially at the
periphery. They are available in 8× or 10× from tool suppliers, but are
rather pricey, around $100. You can save considerably at
To the left is a 40× microscope, with a .02mm reticle and its own light source. It is hand-held, but despite what one might expect it is easy to position and to focus by hand. It is the one I like best when I want to really see what a cutting edge is like . . .
To the right is a 2.25× clip-on style magnifier. It is light weight, so I keep it on my working glasses practically all the time and forget about it. Its power is borderline for looking at a cutting edge, but just right to see what is happening to a surface while you scrape it, and . . . is ideal for pinpointing.
PRECISION SQUARES . . .
LARGER SURFACE PLATE . . .
STRAIGHT EDGE. . . .
4. MASTERING THE STROKES. As you begin this new adventure, keep in mind that it's ok if you make mistakes, you won't break anything. If you make a stroke that is to deep, or too shallow, or out of place, if you miss a marked spot or go over the same spot twice or even several times in the same scraping pass, and almost anything else you might do wrong unintentionally with your scraper, will get corrected automatically in the following passes . . .
The three things that most determine what a stroke will be like are the angle of the scraper, the down pressure on the tip, and the forward movement of the handle . . . Experiment with them, one at a time, and see how each affects the stroke . . . keep in mind that at this point you are not learning how to do things so you can begin to scrape. You are scraping already. And you are learning how to become better and better at it.
. . . look for the stroke depth and length, and the rhythm, that work best for you to remove material efficiently in a sustained way. It is somewhat like finding the best gear for a given terrain when you ride a bicycle . . .
5. EVALUATING THE SURFACE.
IMO it is important to get a sense, even very approximate, for the
size of a scraping project before you start it. Even an extensive
project, if you know what to expect, is likely to turn into a positive
experience, and after you finish it you'll want to do more. The reverse
is true if you find yourself bogged down in the middle of it
unexpectedly. I believe that the stories that come up at times
about the difficulties of scraping, often come from those who found
themselves in the middle of a job, and then discovered that it was much
larger than they thought it would be when they started it . . .
many useful and interesting projects, such as the examples
presented further on, can be happily completed in a few weeks at the most, and
some in less than a weekend.
6. THE MARKING-SCRAPING CYCLE. Now the real fun begins. It goes like this: the surface plate is covered, or "loaded" with a very thin layer of High-Spot blue. Then the surface to be scraped, cleaned of any swarf or other foreign particles, is laid over the plate and moved back and forth. The high areas rub against the plate, and become marked in blue. Those areas are then scraped, and the surface has become a little flatter. That is a complete scraping cycle. It is repeated as needed, until the surface has become as flat as you want to make it. Let us look in some detail at how this is done . . .
As you go on scraping, the total area of the marked surface will steadily increase, and the clear portion will shrink. The surface is becoming flatter, and the distance between the peaks and the valleys is being steadily reduced . . . The proper amount of High-Spot Blue on the plate, and the marking that gives the most useful information, depends on how far along the surface is.
At left both images show the same surface, and the only difference is the amount of High-Spot Blue that was used. At this early stage all the area that marked in the lower image will need to be scraped down, so that between the two, that heavier marking is the more useful one. The right photo also shows the same surface in both images, again marked with different amounts of High-Spot Blue. But here the surface is further along, and the focus of the scraping will be more on the individual spots, so the lighter marking will give the most useful information . . .
8. THOSE DARN EDGES.
When you are scraping a surface, and you reach its right or left edge,
the scraper is prone to tip over, dig into the edge at a sharp angle,
and make a %ζ#@$ gouge . . .   This is a
well known predicament, that can happen occasionally even to
experienced Scrapers. . .
9. WHEN YOUR BODY COMPLAINS . . . If done intensively or for long stretches of time, scraping can be quite physical, just like pushing a hacksaw or a wood plane, or pedaling a bicycle, or swimming, or many other physical activities. And especially so at the beginning, because you are asking your body to . . . bring into play a new or little used set of muscles. They will strengthen with practice, but will complain until they do . . .
Take time to develop the best working position for scraping . . . You may find also that there is one height that works best for scraping,
and a different height to see best what happens when you scrape . . .
10. STABILIZING THE SURFACE. As we continue to lower the high spots of a surface with successive scraping passes, we expect it to become increasingly flat. Most of the time that is indeed the case. But sometimes it is not. It can happen that while we are happily marking and scraping and minding our own business, things begin to feel strange . . . and the trend of successive markings makes less and less sense . . . If that seems to be happening, and you know that your marking technique is reliable, suspect instability . . .
Instability is a condition that can occur when . . . the surface can wobble or roll on the plate, by very small amounts of course . . . so different sets of high spots will touch the plate, and different marking patterns will result.
. . . once the approximate location of the convexity in the surface you are scraping is determined, it
is a simple matter to remove it . . . However if the curvature is
very small, as is often the case, it . . . may well remain undetected. This is especially likely to happen . . .
when the scraping is nearing the end and the quantities involved are very small.
We may end up assuming that the surface is flat, and consider it finished, when
in fact it is still convex . . .
11. TESTING FOR INSTABILITY. There are various tests to check for instability, and . . . to locate the position of the convexity causing it. A number of them are described below . . .
SPIN TEST . . . Place the work on a dry surface plate. Push it gently sideways, near one end. It will rotate (spin) around a point, or pivot. Note the location of that point. Repeat, pushing it in the same direction, but from the other end. The rotation will be centered at a different point. If there is no convexity, the two pivots will be somewhat distant from each other, and towards the ends of the surface. If there is convexity, the pivots will be closer, or even coincide . . .
SIGHTING (LIGHT TEST) . . .
CHANGING THE POINT OF PRESSURE. . . . To check for twist, especially with a long and narrow surface, mark while applying pressure close to one end. The opposite end may mark lighter, but the marking should be uniform across its width. If marking is lighter on one side, as seen in the image, twist is present.
SHIMS . . .
TEST INDICATOR . . .
SENSITIVE ADJUSTABLE LEVEL . . .
SCRAPING TEST . . .
12. MAKING SENSITIVE ADJUSTABLE LEVELS. I made this adjustable interface several years back, when I was aligning my Minilathe, to check for twist and vertical bow of the bedways. I measured twist by placing it on the saddle as shown, and then moving the saddle along the length of the bedways. I measured bow in the same way, but with the axis of the level parallel to the bedways. Since then I found it useful for more and more applications, including checking for convexity of a surface being scraped.
. . . the bottom rests on the work by means of three adjustable threaded rods. The top element pivots at one end . . .   and is supported at the other end by a micrometer head, located by a steel ball, by means of which the bubble is centered in the vial. In this way the level can be used on any surface, whether or not it is leveled, or flat.
Despite the informal construction and the use of wood, this configuration turned out to be quite stable. With the level shown (Kinex .02/1000) . . .   drift is less than 1/2 division over several weeks . . .   you can build such an interface in a weekend or so.
After I started using it I kept coming up with more and more uses for it, however its bulk was often a limitation, or at least a nuisance. So I decided to make a more compact, more versatile, and more "graceful" version of it, using just a replacement vial, which is small and inexpensive . . .
The photo at right shows the complete level, which I made as small as practical. The footprint is 19×120 mm, it weights 550 gr., and can be adjusted over a range of ±10%. The little cross level is removable, and I usually keep it off. The bottom is accurately scraped and pinpointed, and has two holes for installing accessories, such as the three points attachment for using it with irregular surfaces, shown at the top. The micrometer head has a vernier graduated to .001mm, which corresponds to a bubble displacement of 1/4 division, and a slope change of 1/100,000, or .01mm per meter. Resettability and lag are less than one division of the micrometer vernier.
I "made" the micrometer head by sawing off the bow of a micrometer obtained from ebay. Sawing off the bow of a micrometer can be a very satisfying experience, so if you are going to do that (haven't you always wanted to?), be sure to take time to savor it. Over time I also made a range of accessories to go with the level, which enables it to do pretty much all the things a conventional sensitive ("Master Precision") level can do, and some that it can't.
The construction can be simplified considerably with the minimalistic version shown below, which can be built easily, and in a few days. It is fully adequate for many applications where good sensitivity is needed but high resettability or precise angular measurements are not. Instead of a micrometer head resting on a sphere, use a lowly M3 thumb screw. Make a hole tapped M3 at one end of the top element, then machine the thumb screw with a head of 1/2" to 5/8" and thread it . . . adjusting the opening of the dye for free movement but minimum clearance in the hole. Round and polish the tip . . . Then drill a corresponding hole in the bottom element with a #1 center drill to receive the screw . . . At the other end . . . use a length of 1/8" drill rod, relieved in the middle so that only the ends will make contact with the bottom. Use a spring to hold the top and bottom elements together . . . fasten the vial with two clamps made of piano wire. Make the main components out of steel . . . with provision for thermally insulated handles. As the last step, do a careful scraping job of the bottm. The total cost of the level will be something like $30, or around $75 for the micrometer version if you get a used micrometer from ebay . . .
13. MAKING AND USING TEMPLATES. When scraping long and narrow surfaces, for example a dovetail, or the bedways of a lathe, marking with a surface plate is usually not possible or practical. When that is the case, the marking is usually done with a template of the appropriate shape, scraped flat. The template can be shop made, a practice I recommend . . .
14. SCRAPING DOVETAILS. Fundamentally, scraping dovetails is the same as scraping other surfaces. However because of the space constraints, it can be somewhat awkward, especially until you get used to it . . . you will find here a detailed description of how to scrape the dovetails of a saddle and cross slide. It will show as well how to scrape dovetails when different surfaces need to be aligned with each other, or with external elements.
15. THE THREE PHASES OF SCRAPING. Scraping is a process of gradually transforming a surface that is rough and irregular into one that is uniformly flat. The way we do this changes along the way, so it is convenient to think about scraping as having three distinct phases. But each phase blends smoothly into the next, and the change happen gradually, with considerable overlap. The change depends on how the distance between the peaks and the valleys of the surface compares with the depth of a stroke . . .
FIRST PHASE . . .
SECOND PHASE . . .
THIRD PHASE. In the first and second phase all surfaces are scraped in more or less the same way, no matter what their eventual function is going to be . . .   In the third phase . . .   the way we will proceed will be different according to what the surface is going to be used for . . .
16. FIXED AND SLIDING SURFACES.
SLIDING SURFACES . . .   must have sufficient bearing spots to distribute evenly the pressure on the surfaces, and to minimize wear . . . Usually they need to be aligned also with other surfaces, or with other mechanical elements . . . the two sides of a way must be mated carefully, so as to be in good contact with each other along the full range of movement . . .   There must be lubrication between the surfaces . . .   To delay long term wear . . .   finish slightly higher the areas where you expect wear to be greater. By doing this in a judicious amount, the initial wear will make the surfaces more rather than less even, and you'll get more mileage from the machine. To the right is the bottom of a lathe cross slide, scraped in this way . . .
17. HIGH GRADE SURFACES. When scraping precision tools, a high degree of flatness and a substantial number of contact spots per square inch is often called for. 25 spots, as shown in the photos, should be the minimum, 30 to 35 preferable, and 40 spots or even more . . . may be appropriate for small surfaces, when very high accuracy is required. Also, it may be desirable to finish the surface with a pattern that will show any wear or damage that may occur in use . . .
As the surface becomes increasingly refined, you may notice that when you slide it on a dry surface plate or other very flat surface, it may tend to stick to it. That is the sign of a job well done. So the first time it happens, make sure to pat yourself on the back . . .
18. SCRAPING HARDENED STEEL. Yes, it can be done, with material hardened up to 50-55 HRc, and using carbide scrapers of course . . .   Your strokes won't go nearly as deep as you are used to and scraping will go slower, but in most cases the material you'll need to remove, for example to improve the accuracy of ground and hardened measuring tools, will be much less. As an example, the top image to the right is the short side of a 60° template, actual size 7mm × 48mm, showing the first scraping pass. The vertical lines are the grinding marks. Below it is the same surface, after the third scraping pass. The grinding marks, although not completely gone, are now barely visible. The bottom image is the long side of a hardened 45° template, which measures 7mm × 110mm, scraped to more than 40 point per inch . . . The very small black spots . . . are pinpoints.
19. THE FINAL STROKES. . . . If you were to carefully build or recondition a machine for its best performance, at the end you would want to give it a good paint job, and do anything else that will make it look as it rightly deserves. In the same way, an esthetically appealing scraped surface is the appropriate conclusion of a good scraping job. And with such a surface, it will also be easier to see any ongoing wear, and any possible abuse . . .
I have seen various opinions and recommendations, at times contradictory, about how to finish a scraped surface. In Europe in the 50's, when my exposure to "formal" scraping came about, there was a small number of Master Scrapers who kept in touch with one another, and once in a while got together for friendly discussions, competitions, and exchange of information about various aspects of scraping. How to finish a surface was always a hot topic. Several of them had developed unique scraping patterns, which they used as signature on their best work. They could look at such a surface, and tell at a glimpse whose job it was.
Over time I have got to be somewhat good at it, if I do say so myself, although nothing like the kind of things I have seen them do. Here is one of my more successful efforts. It is a small V-block I made to check the 'V' bedway of my Minilathe when I was setting it up, and I scraped it as precisely as I could manage . . .
In this style, which is the easiest and the one I would recommend, the last two scraping passes of the finished surface . . . will scatter light in two different ways. It is that difference which most of all brings about the esthetic appeal of the finished surface . . .   when you look at it you sense the pattern, but without actually being able to see it distinctly. That makes you want to "find" it, and to do so you end up getting more and more into it, feeling all the time that you are getting closer, without ever finding it, but with an ever growing esthetic experience.
20. MARKING TECHNIQUES. We have seen earlier how . . .   the way a surface is marked can bring out or obscure the information that will be most useful at the time . . . The photos below, enlarged 2×, show once again one surface marked in two different ways. . . the boundaries of the marked areas are similar, and the difference is that in the left image the intensity of the marking is rather uniform throughout, while in the right image it has a pronounced gradation, which shows at a glance and in detail the landscape within the marked areas. Note that the gradation is anti-intuitive. It is lightest at the peaks, darker in the somewhat lower areas surrounding them, and darkest of all where the marked areas are lowest, just before the marking suddenly stops . . .
LIGHTING. As long as the marking pattern has enough contrast . . .
most light sources will do. But as the scraping progresses . . .   you will need to
continually cut back the amount of the High-Spot Blue on the plate . . .
the contrast of the marked pattern will
decrease, and the marked spots will become . . .   increasingly difficult to see.
. . . There are two things you can control to make the marked spots stand out best. The orientation of the light source, and the color of the light . . .
The markings are blue, therefore orange light . . .   will produce the best contrast, and very long life incandescent bulbs, which are an orangy (warm) red, will work best. Normal incandescents are pretty good, and fluorescents, even the warm kind, are the worst . . .
These measures will extend the useful range of the High-Spot Blue, but eventually the contrast it can provide will have reached its limit . . .   In order to go further, there are other marking techniques that can be used, such as pinpointing, or marking by evaporation.
21. SPREADING IT THIN. I have seen a number of suggestions about how to spread High-Spot Blue on a surface plate, from using printing rollers (brayers), which are quite popular, to doing it with your bare hands. I use shop made fabric pads, which I find most convenient to spread High-Spot Blue in a controlled and uniform way down to very thin layers, thinner than is feasible with rollers, and keep my hands reasonably clean. There are warnings about fabric dropping fibers on the surface plate, and causing trouble. I had to conclude that they must be urban legends, as I have used fabric pads for almost all of my scraping, and I never had problems with them shedding fibers, even on occasions when, after much use, the fabric wore completely through.
A quick way of making such a pad was described earlier. A more "formal" version is described here. It is simple to make, and has a flat bottom and a handle, which makes it easier to use and will keep your hands cleanest . . .
It is a very precise technique, which can take scraping considerably
beyond the limitations of High-Spot Blue . . .
in the later stages of scraping a high grade surface
. . . peaks will rub directly on the surface plate, and become burnished and somewhat mirror-like . . .   Like mirrors they will reflect light without scattering it, so if you look at the surface with a light source at the proper angle, they will appear as very small intensely bright spots. They will be much brighter than the rest of the surface . . . And in the same way, if the light source is at an angle where it is not reflected by the pinpoints, they can appear dark, or even black.
Pinpoints appear naturally, and in increasing number, as the High-Spot Blue nears the end of its range. Therefore you can begin pinpointing while you are still marking with High-Spot Blue, by scraping primarily the top of the high spots that show a pinpoint. After a while you can stop using High-Spot Blue, and shift to pinpointing "formally".
By scraping away just the pinpoints and a little
of their contours, you can bring the surface to a remarkable
degree of flatness and a very high number of bearing spots, well beyond what is
possible with High-Spot Blue
. . . The photos . . .   show what a pinpointed surface is like. The small photo shows the surface in its actual size. The middle photo is enlarged 4×. Pinpoints don't show on the right side, and are less visible on the left, because of the lighting constraints in taking the photo. I had to use a 40" overhead fluorescent oriented to the long dimension of the surface, and the pinpoints that show are the ones that reflected that light, but pinpoints are equally present on the whole surface.
The large photo is the middle section, enlarged 20×. You can see the different amounts of burnishing in pinpoints of different heights, and you can see also in the background the marks of the scraper, which show the last two passes at about 90° from each other . . .
Pinpointing is part of the repertoire of an accomplished Scraper . . .   I believe you will find it interesting and fun to try, and to find out how far you can carry the process. You may have immediate applications for it, or if not it will become part of your "bag of tricks" should a need for it arise in the future.
23. MARKING BY EVAPORATION. . . . With High-Spot Blue, beyond a certain point the contrast becomes minimal, the high spots are increasingly difficult to distinguish, and the intermediate values of depth are practically absent. A surface marked by evaporation with a light color pigment shows better contrast, and a marking pattern which is easier to read. It also eliminates the annoying scattered light of the scraper marks, and therefore makes it possible to utilize fully the contrast that is available.
The surface shown in the photos at right, enlarged 2×, was scraped close to the limit of the High-Spot Blue, and then was marked both ways. With High-Spot Blue, marking had to be done keeping the blue on the plate to a minimum, with little rubbing, and light pressure, to avoid filling in the lower areas. The marked spots are difficult to see, the contrast is quite low, and filling in almost happened anyway at the bottom left.
In the surface marked by evaporation the high spots are dark, and the lower areas are orange. The contrast is considerable, the pattern of marking is well defined and includes a good range of intermediate values, and the profile of the landscape between peaks and valleys is readily apparent. Note for example the small high spot inside the circle on the upper right. When marked with High-Spot Blue it is barely noticeable, with hardly any detail. When marked by evaporation it stands out clearly with good resolution, and the details of the landscape around the peak are easy to see.
Pigments are insoluble and often intensely colored substances. They are the colors that artists use when they mix their own oil and acrylic paints, so are widely available from artist suppliers, ground in very fine particles. When marking by evaporation, the function of the pigment is not to mark the high spots, but to fill in and dull the lower areas, so that the high spots, which after marking will end up bare, can be more easily seen. Therefore the surface plate is kept dry, and a light colored pigment in an alcohol suspension is applied to the surface being scraped. Then the alcohol is allowed to evaporate, and a thin layer of pigment remains . . .
25. Making a large square
Scraping an 8"×12" square, accurately, in a weekend.
26. Dovetail template
The sequence of refinishing the working surface of an 8" long dovetail template.
27. From beginning to end.
Scraping a surface to a high degree of flatness, and a considerable number of bearing spots. The details of the scraping sequence, which is typical for many surfaces, are described and illustrated.
28. Self-generating reference square
A 4"×5½" square is scraped to an accuracy of better than ± .002mm/.00008", without using external references.
29. Upgrading a 45° hardened angle template
Example of scraping a hardened and ground surface, typical of many precision measuring tools.
30. Squaring the ways of a saddle
The top and bottom ways of a lathe saddle are scraped square with each other, to the desired degree of accuracy, without doing any machining with the lathe, whether or not the lathe is assembled, and using a minimum of measuring tools.
31. Fast and accurate setting of the lathe compound angle
After having scraped three surfaces, the compound can be set to zero, or to any other angle, with precision and in a few moments. Includes a Minilathe modification to immediately access the compound bolts, at any position of the slide.
32. Scraping a saddle and a cross slide
The sequence of scraping all the sliding surfaces of a lathe saddle and cross slide, in proper relation to one another and to the bedways, and using a minimum of measuring tools, is described in detail, with many photos and comments.