Time: 7:37:21 PM
Topic: leaving tracks as you sand or polish
jan 24 2007
Note on the problem of leaving sanding or polishing traces (tool marks):
Background: Sanding or polishing is a process of repeatedly moving progressively finer abrasive surfaces over an object surface until the object’s surface becomes smooth. Each step uses a finer grit abrasive which is rough enough to remove the previous abrasive abrasions. A problem occurs when the abrasive surface makes uneven contact with the object and leaves tracks rougher than the next finer abrasive surface can remove.
[Note: sanding (“lapping”) and polishing are the same process, where polishing is just using finer grits. Purpose of sanding is to remove grinding or other tool marks. The purpose of polishing is to achieve a optically smooth finish.]
The cardinal rules I had developed for myself were to polish at least two orthogonal directions and to polish on the bias of the curve. My old style was to make many fast strokes that eventually all overlap, sweeping around the contour. With this piece I thought I was leaving too many stroke marks doing it that way. So, I switched to a slow, vertical, and overlapping series of strokes. Off course my vertical only strokes violated both principles and thus I am left with big polishing tracks. [later I realized I was not overlapping properly, using the edge of the pad instead of the inner area, and not fairing with a full contact pad before going to next finer grit.]
The problem illustrated. This picture shows where I used a 4" angle polisher to polish a basalt column dome up to 400 grit with wet diamond 4" pads. Following that I hand sanded parts of the surface with a rougher diamond pad, 200 grit, in a horizontal direction so that the ridges and valleys of the polishing tracks are visible.
Basalt polishes to a deep black at about 400 grit and finer. It's scratch color is light grey.
Before I sanded over the polish, I could see splotches of foggy grey. I didn’t realize they would turn out to be relatively deep polishing tracks. Knowing now that they are there, I can partially make out these tracks. This is where one experiences how to "see" not unlike early natives who had to learn to see subtle tracks and spoors of animals in the wild.
The ridges and valleys left by the vertical stroke are visible as light and dark stripes. The ridge gets the most contact with the hand pad and is scratched grey. The hand pad doesn't scratch the bottom of the valley so the valley remains darker from the 400 grit polish.
(What I am not sure of is how much of this ridge and valley artifact is from previous grits. I went from 50, 100, 200, and 400. I did some hand sanding at earlier stages to smooth out these same kind of tracks.)
Obviously I have to do something different. So, I used a backing pad with a polishing pad to just hand rub it across the contour to get the feel of how the pad contacted to stone surface. I suddenly realized that I could get much more flat contact by holding the pad almost flat so that it contacted from the center out.
What I now have to find out is how it will work while spinning - for example so that it doesn't chatter or bounce. In the morning I will experiment to see how to get the flattest contact with the domed surface.
(two other variables now, new pads and new angle polisher)
technique to note: using angle grinder/polisher on a convex surface
strokes that transverse the convex contour (a ball or dome)
using wrist and arm motions
minimize or control faceting tracks
pressure
angle of contact between pad surface and stone surface for smooth abrasion
rotating the grinder to keep the pad at a constant angle to the surface throughout the stroke.
Day 1.
Best stroke pattern. I now have gone back and tried the many fast strokes in multiple directions. It does work better. What I still don’t know is how to be sure I have the maximum surface contact. I am using a full sized hard rubber backing pad (matches size of polishing pad. On a dome shape I don’t believe that an undersized pad helps, but it remains to be verified.
Revealing and Seeing. Now my curiosity is piqued and I am determined to figure out how to polish the dome shape. I can now see the polishing tracks through the water. [#fig needed#] Normally I complained about how water hides any scratch pattern above 100 grit. However, when I look for the tell tale grey streaks, I can just barely make them out. The weather is overcast and misty so I can’t easily switch to dry sanding to verify where the sanding tracks are. It appears that it is best to lightly hand sand with the same grit to reveal the unevenness. The pad must be flexible and thin enough to conform to the surface curvature, but doesn’t conform to the small ripples and irregularities. Only a few strokes are necessary to reveal the tool tracks. More strokes may be necessary to remove the unevenness.
The new pads I have don’t chatter at all. The new polisher is set at about 4000 rpm and runs a little faster than the old polisher. The extra speed seems to help.
Day 2. The second day studying how to manage polishing tracks on a dome shape brought more uncertainty.
• - basic problem we are addressing is how to polish with a series of finer grits, without leaving traces of the coarser grits. For example the object is polished first with 100 grit pad, then 200, 400, 800. 800 grit leaves a very glossy surface. A glassy surface is achieved by continuing to double the grit up to a few thousand. The main traces that are left are facets the width of the abrasive contact surface and the length of the stroke. Facets are flat when a rigid abrasive pad is used or slightly concave grooves when the pad has flex. A series of overlapping facets creates a series of ridges and valleys the length of the stroke. Typically a second pass is made where the strokes cross the previous strokes in a cross hatch pattern. The facets become squares or circles where the strokes cross. With skill these facets are removed with each successive stage of finer grits. However, in practice, tracks are left behind. A track is left when the polishing overlap was not exact. A track is left when uneven pressure creates a depression to deep for the next stage of grit to remove.
• hand sanding with a thin pad has the best contact with the convex surface and seems to be a necessary step.
• I still am not sure where the sweet spot on the polishing pad is for domed surfaces that are convex in all directions.
• the area of contact seems to be less than a half inch wide. That means you advance the next stroke less than a half inch.
• fatigue plays an important role. It is easily to start leaving deep tracks after fatigue sets in.
• still not sure when to use fewer slow strokes that slightly overlap or many fast strokes with lots of overlap.
• sweeping stroke versus swirling stroke. A swirling motion makes it easily to advance slowly and evenly. But the swirling motion makes it much harder to follow long contours evenly. A sweeping stroke makes long contours more even.
• very light pressure is used for polishing, less than the 5 pound weight of the polisher.
• backing pad can be on size or under size. For concave surfaces the undersized backing pad allows the thinner polishing pad bend and conform to the cavity. For convex surfaces, the situation is not so clear. It seems that its best to keep the middle of the pad in contact with a convex surface. But that requires the polisher to be angled just right and adjust exactly as the stroke transverses the convex surface. Using an undersized pad and a sharp angle sometimes seems to work because the polishing pad will flex to match the surface making up for slight errors of the angle of the polisher. Using the outside edge of the pad can leave a deep track.
• bright sunshine helps reveal polishing tracks even wet and with 200-400 grit pad. Sunshine makes it more obvious that polishing is done fairly blindly, that is, it is always hard to see the results as they are made.
Day 3. Third day trying to figure out how to polish with an angle grinder.
Somehow I managed to get the basalt dome up to a good 400 grit polish today [or so I thought at the time!].
The foam pads tended to chatter today, so I switched to hard rubber pad which ran very smooth. Even that pad would chatter if I held it wrong. Couldn’t quite figure out what the pattern was.
I had to do a lot of hand sanding today to smooth out tracks left from previous grits. A lot of my problem is from fatigue and not holding the polisher consistently. Also, visibility is usually terrible and doesn’t let you see your immediate results. Sometimes the sun is just right and you can see closely what the polishing pad is doing.
There is a big difference between diamond pad and silicon carbide sandpaper grit sizing. 220 grit sandpaper seemed to be close to the 400 diamond pad.
I was surprised to see how worn out my diamond hand pads were. They wore out a lot more than I realized, but I unconsciously applied more pressure. It is almost as if the first 50% effectiveness wears out within an hour or two and then the last 50% wears slowly over days and weeks. Lesson: spend more on new pads!
Day 4. I ended the day yesterday depressed about the polishing results. Today after bumming around half the day, I began hand polishing with my two 400 grit pads. Before long I was really getting into it. The pads did their jobs if I was patient. And the sun was bright so I could see the surface at that resolution clearly. I was able to clear up all the gouging tracks from earlier grits [but later I discovered differently]. And I was able to do the edges fairly well.
Keep in mind day 4 only starts from my curiosity being piqued. My total labor hours are in the range of 100 to 200 hours over several weeks from November 2006 when I started, about 9 weeks ago.
Issues and parameters to explore:
• straight stroke versus swirling stroke by hand
• rotary pad (angle polisher) versus linear stroking pad (belt sander)
• pad flex and compressibility to allow it to conform to the surface or to impose a flat surface
• short range and long range roughness
• various tool marks from edge gouging, surface unevenness and tool defects, pressure applied unevenly, end or begining of stroke differences
• conceptual symmetry - along various axes, consistency of curve type and edge sharpness, etc., various patterns that are set up. Roughness is all that detracts from that symmetry. However, part of the symmetry the piece may be intending is patterned roughness. In that case one has to decide how to classify roughness or otherwise undesirable features.
• fatigue and impatience leading to careless tool handling
• differences among the various abrasive technologies -
• buffing wheel and loose abrasive versus pad
• resin/wet versus resin/dry (diamonds embedded in resin)
• bonded diamond (diamond pad) versus bonded silicon carbide (sandpaper)
• dynamics such as chatter, speed, pressure, water, natural radius of fingers, wrist, arm, etc (skeletal motion)
• match between shape of abrasive pad and surface (e.g. file over cylinder, etc.)
• develop a “lay” in which the stroke and scratch lines are oriented in the same direction, as opposed to strokes going in various directions where the scratch lines cross. Then change the direction of the lay to cross hatch the surface.
• abrasive bonding technology - nickel bonded, resin embedded
• wet versus dry sanding
• at finer surfaces: contaminating grits are a worse problem. This is probably the reason diamond pads at 800+ contain well spaced nodes of abrasive.
• fatigue ruins your best intentions, you overlap less and use more irregular strokes, irregular pressure
• very hard to see the last few remains of scratches and facets so they show up, as unwelcome surprise, after you complete the next finer grit. Moral, examine twice before moving to the next grit.
• hard to see wet. use crayon or something waterproof to mark the surface and then watch how the marks get removed.
Below is a picture of the basalt column dome with an almost complete 400 grit finish. There were still a few minor tracks, foggy spots, and deep scratches that I fixed before dark.
Day 5 of the polishing study.
polished about 20% of the 400 grit surface to 800 grit. worked a couple hours for one square foot. Had to hand sand to get the pad streaks out. Used water and an 800 grit resin hand pad.
It is possible to hear how smooth the surface is by listening to the fuzzy buzz as you hand polish. A fresh area has a loud full sound. The sound gradually softens and thins out as the surface matches the pad. Also, it is interesting to note how the sound can become words of a repeating phrase in your mind.
This study is becoming more generally a lesson in how to abrade a convex surface using a spinning polishing disk and hand pads. The fact that the disk doesn’t conform to the shape is the main issue, made worse by lack of visibility because of the water.
Discovered a big area where there were deep polishing tracks from previous grits. These tracks only show up when you hand sand over them with a stiff flexible hand pad. Just goes to show how subtle these tracks can be.
Day 6. Today I just threw caution to the wind and used the polisher to bring the whole piece to 800. Then I hand polished a corner to be sure there weren’t any weird artifacts. It seemed to work fairly well.
However, I did notice one general mistake while using the hand pad. It’s abrasive matrix is arranged in widely spaced rows of knobs containing the abrasive. When you pass over the same area several times, your strokes run parallel with varying overlap or spacing. The individual rows of abrasive leave tracks with wide parallel spacing like a rake pattern.
Day 7. Today was the last day for awhile. I hand sanded the piece to 800 grit which provides just the right among of gloss and depth. The mineral matrix is visible clearly and shows how the mineral compound varies slightly in different parts of the stone.
After a few hours of very slow going I began to realize I was learning how to sand at that very fine stage. Ridges and valleys, oh how I hate thee! One big problem is strokes that fail to over lap, leaving these ridges and valleys. Since there is so little contact between the flat disk and the convex surface it is easy to progress too fast. The strokes were about 4 inches in length while the contact area was about 1/4 to 1/2 inch. That means you have to move ahead only a 1/4 inch or less at a time, or alternately, come back and redo previous stroke areas. When you move ahead too far and never come back to overlap the area, the next level of polish shows this as ridges and valleys. Pressure variations make it worse.
Fairing is a key step in the process. The trick is to hand sand and use a stiff but flexible backing pad to “fair” the entire curve before moving to the next stage. That way most of the polishing pad is in contact with the surface and the stiffness distributes the pressure evenly over a large area - thus the curve takes on a long range evenness. Another technique is to use flat greenstone blocks (silicon carbide) or flat diamond files. However, they are so rigid that great care must be taken to use long sweeping strokes that perfectly overlap.
If I had been using a belt sander, all this would have been done without much thinking about it. Note the opposite problem caused by fairing is that the curve can be made too shallow since the pressure is towards straightening the curve. You have to remind yourself to not loose the feature while trying to smooth the surface.
Perhaps I have been fairing at too fine a grit, in other words, after the fact. That is, it is important to fair the surface before moving on the the next grit. Even though, it is easier to see the tool marks at the next higher grit. The problem is that some tracks are too deep to take out at the current grit. Pay me now or pay me later.
In all this technical mess, I am reminded how incredible a hand polished basalt surface is.
Basalt Column with weathered surface, naturally split in the middle.
Circular area ground smooth and polished.
Design by mja, finished by bb.
January 2007
Below is the piece in early work during a cold rainy winter!.
