Flattening Cast Iron Tops by Planing, Grinding or Scraping
Many old arn surfaces were originally factory finished by metal planers. Over time, the pressure to reduce costs led to the phasing out of such processes by American producers. The industry instead turned to Blanchard grinding for surfacing. Even Oliver was grinding like a mad man at the end. Now, more and more, we are seeing large flat bed CNC type grinders similar to the Mattison surface grinder used by Grizzly.
Yet several makers in Europe still refuse to give it up; Martin, Hofmann, Knapp, Logosol, etc. The list goes on. The reason is clear – When it comes down to preparing cast iron surfaces for woodworking machines, the planer and its cousin, the metal shaper were/are the cat's meow, at least if you are getting the surface flat with a machine.
Does My Machine’s Surface REALLY Need Redoing?
OWWM’ers with an out-of-flat cast iron surface, such as a tablesaw, jointer, etc. are well advised to try other simpler solutions before embarking on resurfacing.
The first thing is to understand if the amount of out-of-flat is acceptable in the context of how the machine is to be used.
For example, crowning on a tablesaw top is generally considered to be less of a problem than dishing. Bringing other tolerances in the machine back into spec may do more to improve the machine’s operation than a regrind would.
Shims can often be successfully used to slightly raise a ‘droopy’ corner. Lifting up or weighing down a corner to try and ‘spring’ it back into shape can also be tried. Care must be exercised to avoid cracking the cast iron however, and the table may want to gradually return to its previous shape when the pressure is removed.
Your local machine shop tolerances may be greater than the amount of out-of-flatness presently in your surface. Clearly, nothing is to be gained in this instance. Things can only get worse!
If after everything else that you can think of has been tried, resurfacing seems like the only option, then you will need to seek out a competent shop with a machine capable of handling your top.
If you are very lucky, there will be a shop with a metal planer nearby. More likely, you will have to settle for Blanchard grinding. Searching the forums is a good place to start, or the yellow pages for grinding in the nearest big city. Look for references to precision grinding. Car racing or engine rebuilding shops can sometimes be helpful.
Make sure that the shop can handle the physical size of your machine. If it is a jointer, and it is not possible to shim the ways, it will need to be reground as a single unit.
Planing
There is general agreement among OWWM’ers that the best surface for accuracy and friction free surfacing is a planed surface off a broad nosed finish tool. Planing tops is not cheap. It is the most complex and time consuming way of finishing a top surface. That is why the process was abandoned by American manufacturers many years ago.
And, for OWWM’ers, it is getting harder and harder to find any shops that still have a metal planer in the back and a guy who knows how to run it right. These dinosaurs are really that.... dinosaurs! A typical planer that can do an Oliver 88 table in one setup will be 20 feet long and weigh in at about 25,000 to 75,000 pounds! They sure can do accurate stress free work though.
A metal planer leaves a dull, gray, corduroy finish known as tigerstripes. The stripes are left from the width of the planer cutter when it was run through. Even after 50 years of burnishing use, these stripes are often still visible and an indication of how much surface wear there has been.
Many older machines with planed tops are still very serviceable, and the planer marks are often still there to this day. This is likely due to the fact that the planer marks left on the surface reduce the friction on material passing over it significantly, thus the amount of wear is on the surface is correspondingly reduced. Such tables are not dead flat but rather scalloped, and over time the ridges of the scallops begin to wear. But since these ridges account for only a small percentage of the table's surface area, the majority of the surface does not get worn at all. Small amounts of very fine sawdust and wax from polishing float in these tiny recesses between scallops, also reducing the friction. The result is a table top that wears similar to the hydrostatic ways of a milling machine with reduced wear and reduced operator effort. This translates to significantly reduced wear over the long haul; thus, the high number of older machines whose tables are in pretty darn good shape.
Metal planed surfaces are traditional, practical, long lasting, and cut down on the stiction. They are also known for being damn accurate! Even today, it a takes half million dollar CNC grinder to achieve the same level of accuracy as a metal planer built in 1920 can and did.
Metal shapers use a point tool and the tool moves to and fro in a reciprocating motion. Very little loading is put into the parent metal, thus little deformation due to mechanical load or heat.
Grinding
There are three typical methods of surface grinding:.
- Reciprocating table, vertical spindle machines (think large cup wheel, or segmented), which are less common.
- Reciprocating table, horizontal spindle machines which are what most people think of as a "surface grinder". Given equal machines and operator quality, the reciprocating table horizontal spindle would be best quality for a long flat surface. The grinding pattern is more appealing, being almost invisible, but parallel to the long dimension. However, it is slower, and hence generally more costly.
Reciprocating table grinders wear so the centers of the ways go hollow. On large surfaces, the result can be dome shaped or potato chip shaped surfaces, albeit the slopes measured in small increments.
- Blanchard grinding, which is the most common. Blanchard grinding rotates the table (a magnetic chuck) under a very large segmented wheel. Both the work piece in the chuck, and the grinding wheel rotate. The typical swirl pattern you see on many new tools is evidence of the Blanchard grinding process.
In any case, quality is going to be primarily dependent on operator skill, awareness, and interest; followed closely by the condition of the machine. Lots of big old grinding machines have serious wear on them.
Blanchard Grinding
The Blanchard method creates a somewhat coarse swirl pattern about a focal center on the part (which may be off center to the part, depending how it was mounted to the chuck). Well set up Blanchards grind very flat. They are the standard today because they are also very fast, and when set up well grind very cool.
Delta used a Blanchard grinder, as many Asian companies do today, but Delta left the surface very coarse for two reasons. Because it was very cheap to do, and because the swirl marks help in breaking up stiction (a suction force that makes it harder to move something through).
A table saw top is a good job for a Blanchard, being more or less square, and having no raised edges. Tablesaw tops have a particular hazard when ground however, the miter gage bar will likely ride proud of the miter slot after the grinding process removes material from the top. The next step would then have to be a milling machine to re-cut the miter slot depths. THAT would take more time to do properly than the grinding.
Blanchard grinding can only be used on completely flat pieces. Planer tables with their raised edges or ‘lips’ are not suitable for Blanchard grinding.
When grinding a jointer, it may be necessary to mount both tables to the base in order to ensure parallelism of the two tabletops. Lock them down as parallel as possible. This is what for example Rockwell did when they ground their jointers at the factory.
The Blanchard grinder must be larger than the maximum diagonal measure of the work piece. Blanchard's are usually referred to by chuck swing. A 6' ft. Blanchard will swing a 6' dia. circle. That would be a medium size Blanchard.
Very Large 300" Blanchard Grinder
Note that if there is minor error in the machine, it will tend to make the ground top (very faintly) torriodal shaped, radiating from wherever the focus is. Blanchards wear so that the top droops forward. This creates a broad hollow trough effect, or a concentric pattern of swales. They can end up not producing perfectly flat work on areas that are large, even though they will do fine with large batches of small parts.
Surface grinding by its nature creates heat. If not done carefully this heat will cause stress on one side of the surface, which can ultimately lead to warping, especially where the cross section varies in thickness or is simply thin to begin with. A highly glazed surface is one indication of excessive heat during the grinding process.
Blanchard Grinding Costs
OWWM’ers relate what they paid for Blanchard grinding their tables. There is quite a variation in pricing, so it clearly pays to shop around for price as well as a knowledgeable shop. Which would you be happier with – a crappy $150 job or a proper $250 job?
- $400 to have two tables from a 16" jointer surface ground. They would have done the Blanchard grinding for slightly less.
- Around $220 to have the two tables and the fence Blanchard ground on an 18" jointer. It included some cleanup on the steel edges and some work to prepare the backs for mounting on the grinders.
- $220 for the two tables and fence Blanchard ground on a 12" Crescent. Then had the same shop blanchard grind a 16' powermatic planer bed {$150}, but they had to grind off the raised lips on each side.
- Less than $100 for the tables on an 8" Powermatic jointer to be blanchard ground.
- $100 to do the top on a PM66, reground by the company that does PM's US made saws and shapers .
- $60 to do the platen on a drum.
- $60 each cash to Blanchard grind both a PM 66 and a Unisaw, both with two wings.
- $60 to Blanchard grind a spare table saw top.
Scraping
Some older woodworking tool surfaces were hand scraped, although this would be more frequently true for ways rather than flat surfaces such as tabletops.
Hand scraping as a process allows a person without huge machine tools to make a worn machine just as accurate and "new" (or better than new) functionally as desired.
The basic idea with scraping is to use reference flats, like a straight edge or surface plates to mark a questionable surface, and then scrape off all the "bad parts" identified by the marking spots left from the "ink" on the reference. Sort of like a dentist touching up a crown. Eventually the surface will conform to the reference, over enough iterations.
Hand scraping is not light work. Depending on the amount of material to be removed, it may take several full days or even longer to do. You have to get the mindset of not fooling around, and shoveling off the chips. It is definitely a metal removal process. The work is hard, tedious, and filthy dirty.
But it is addictive to many because with a few simple tools, it is possible to recreate the basic process of "precision" from nearly first principles right up to whatever level you care to attain. OTOH, for the addicted it is like all the other vices. Two weeks ago you didn't know the process existed. Next week you'll be scraping flat parts with a scraper made out of a dead file. Six months from now, someone reading this will be trying to justify to (or hide from) his wife why he paid $1,200 for a used Biax ELM7 off ebay because "honey, it's less than half the price of a new one and now I can finally rebuild all those machines that have been cluttering up your side of the garage."
Seriously, it’s the thought process that is valuable. It becomes an intriguing mental game to figure out where things are in terms of alignment, where do you want to take them, and then how to get there from here "elegantly" with the simplest set of effective tools.
A couple of good threads showing the progression of a surface scraping project are
here and
here. It really shows how the surface markings change as it gets flatter and flatter.
Wikipedia has a description of hand scraping
here. There is also a link at the bottom of the page to making straight edges, and an interesting piece on ways and scraping. Be cautious, as the article in Wikipedia does not show any scrapers suitable for surface scraping. The ones shown are for scraping plain round bearings such as babbit, etc, for shafts. Even the straight one is too small to use for anything more than fitting small pieces, and it looks awkward (short, no spring, poor end shape & handle).
The one on making straight edges is good in theory, but there may be some issues in practice. The gages referred to as "straight edges" by people who scrape look like the ones in the PA shop article with wide scraped bases. They do not look like knife edge rulers.
The part about using abrasives is more or less true, but it is slower than scraping, by a -lot-. Whitworth essentially settled that question in the 1840's, and laid out some of the issues by which the slower abrasive process also tended to generate its own error set, constantly setting the process back just at the point where things were starting to get flat. He very ably and concisely remarks on and dissects the favor with which abrasives were used because essentially they fooled both the workman and the purchaser by the shiny surface and the apparent amount of sweat expended.
This is an ad for commercial Anderson scrapers, the flat style.
Anderson Scrapers
They also make a tubular style. They are on somewhat expensive, but can be found at flea markets occasionally, or copied. There are from time to time Chinese imports. KBC used to carry them, then stopped. Kitts did the same. Just an example of one style.
Dapra sells a
version that only takes Dapra inserts. Again, pricey but good for information.
Also a
good description of the various power scrapers (Biax)
Most people start out making their own tools, both the scrapers themselves, and, over time the references (straight edges & flats) as need and opportunity arise. Here are some of mine.
Home Made Scrapers
Most people start out by making a scraper out of an old 12 or 14" flat file. The evolution usually goes toward silver soldering a carbide tip on a flat, scraper shaped piece of steel because it is possible to scrape a lot longer between sharpenings. Unfortunately, carbide requires diamond wheels to shape, or at least diamond (hand) hones to keep sharp.
These are blades for a Biax. The cutting edge geometry is the same for hand or power, an obtuse angle (about 93 deg) on each side of the curved cutting edge for cast iron and most other metals. The blade at the top is for steel, such as the RAMCO sander bed. Mild steel is the only material scraped with a positive (acute) cutting angle; even then only by power scraper since it would be nearly impossible to push by hand.
Biax Blades
Here are some basic references, a surface plate, and a 30" straightedge. The surface plate was used to scrape the SE.
A Surface Plate and a 30" Straightedge
Straight edges are available up to about 10', but they get difficult to handle over about 5' long.
A good place to read for days on the subject is the
Practical Machinist site. Search Scraping, Handscraping, Scraper, Hand scraper, Straight edge, Biax, Dapra, etc.