Replies

1. 
   Kevin | Jun 22, 2003 10:05pm | #1

   That's a fairly complex issue you've raised, Wanda. Some people like to pre-raise the grain by applying water first and then lightly sanding the wood after it has dried. That is not strictly necessary. But, it does ensure that the wood pores are open.

   The short answer is that your friend is wrong. 220 grit is the ideal final sanding grit for most species of wood in my experience and that of many, many commercial finishing operations. However, different kinds of wood will accept stain to different degrees. Maple for example is notorious for being difficult to stain. Pine, on the other hand, is exceptionally easy to stain.

   What the stain "bites into" is the wood pores and fibers. As long as the wood pores are open, the wood will accept stain just fine... particularly with pigmented stains. Stains that are mostly or completely colored with dyes will bite into any wood regardless of sanding because the dye will penetrate the wood fibers, unlike pigments. What the sanding does is it gives you a uniformly opened wood surface so that the wood will accept the color uniformly. Most commercially available stains are a mixture of pigments and dyes. So, prepping the wood surface is important to the eventual color.

   A really good and easy way to get a handle on the relationship between sanding and staining is to get a piece of wood and sand the surface uniformly with any grit you choose. Then take a nail or something like that and place a fairly deep scratch across the wood surface. Then on another part of the wood surface use the back of your fingernail or anything smooth and hard and rub it across the wood surface to create a shiny streak that you can see by tilting the wood at an angle to the nearest light source. You'll see the light reflecting more off of the shiny streak where you just dragged your thumbnail. Then take a pigmented stain and stain the wood surface. The deep scratch will stain darker than the rest of the wood surface, while the shiny streak will stain lighter than the rest of the wood surface. The reasons being that the scratch opened up the wood and gave the stain pigment somewhere to lodge and remain after you wiped the stain. The shiny streak sealed off (or closed) the wood pores leaving the stain pigment nowhere to lodge before you wiped it off. The nail scratch would be akin to sanding the wood surface with a really course grit of sandpaper. While the shiny streak would be akin to sanding the wood with a super fine grit of sandpaper.

   Regards,

   Kevin

1. 
   wendy | Jun 23, 2003 04:32am | #2

   Hi Kevin,

   Thank you for answering my question regarding sanding. I will pass that information along to my friend.  I was a bit surprised when my friend said a 220 grit sand paper was too fine. I've always used 220 for the final sanding. especially on pine. I will have to set him straight. LOL

   wanda

2. 
   forestgirl | Jun 23, 2003 07:34am | #3

   Wanda, I've been amazed over the last several months at how adamant different woodworkers are about the "ideal grit" that they have settled on.  They vary wildly, and most people are convinced they are absolutely right, so your friend is not unusual in his statement.

   In practice, the best thing to do is select a couple or 3 grits and make some test boards using your stain of choice.  A higher grit may produce a lighter color with one type of stain on one particular wood, and yet produce a darker color on another wood or with a different stain.

   forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl   ;-)

3. 
   HowardAcheson | Jun 23, 2003 06:53pm | #4

   Sanding wood--hard or soft--beyond 220 does nothing more than burnish the wood making staining difficult. This is particularly true if you are using a pigment stain which sits on the surface and relies on "nooks and crannies" to impart color. Softer more porous woods can be sanded to to 220 but harder less absorbent woods may stain best if only sanded to 150.

   A number of years ago a shop I was involved with did series of adhesion tests with various finishes and sealers. As part of this test we explored adhesion based on sanding grit. We found about the same adhesion up to 180 - 220. Beyond 220 adhesion dropped off due to burnishing of the underlying wood particularly when non-linear machine sanders were used. This was tested on birch panels. We also found that the resulting smoothness of the first coat of finish was not materially affected by the smoothnes of the underlying wood for sandpaper grits between 150 - 220.. Thin finishes such as lacquer could be affected by grits less than 180. The smoothest surface substrate for final finishes was obtained by sanding lightly after the first coat of finish was applied and dry. Which makes the case for a thinned first coat of finish.

   So our conclusion was that sanding beyond 180-220 was a waste of time and could be actually detrimental. But, there was a big appearance affect if the surface was not hand sanded in the direction of the grain using the highest grit used on the sanding machine. A ROS at 180 grit required final hand sanding with the grain for optimum appearence. If not hand sanded, swirl scratches would show. Final hand sanding in the direction of the grain is a must.

   To carry it one step further, sanding at 320-400 grit after the first coat and subsequent coats was the optimum. Any finer and adhesion problems were encountered. No improved appearence was noticed by between coat sanding beyond 400 for varnish. 400 was the sweet spot for thinner finishes.

   I think you will find similar thoughts in the popular finishing books but YMMV.

1. 
   MarkRoderick | Jun 23, 2003 07:50pm | #5

   As many of you might already know, there's a question on this point in the current edition of FWW. Some poor guy sanded all the way through 600, maybe higher, and got terrible results. Chris Minick or one of the other finishing experts replied along the lines above, i.e., don't go higher than 220.

1. 
   Sancho | Jun 23, 2003 09:09pm | #6

   i usually stop at 180 

   Darkworks:  No Guns No Butter squilla and the bling bling.

2. 
   forestgirl | Jun 23, 2003 09:11pm | #7

   While your experiment may indicate a general effect, I think the statement "beyond 220 does nothing more than burnish the wood making staining difficult" is an overstatement.  I, personally, have done stain samples with alder and gel stain wherein the higher grit (320) produced a darker and more pleasing stain than the lower grit (220).  There are any number of skilled woodworkers who do sand beyond 220.  All that sanding may not be necessary, only they know, but I still maintain that the way to determine personally what grit will work, especially when learning about different woods and different stains, is to run some samples boards with 2 or 3 grits.forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl   ;-)

1. 
   Elcoholic | Jun 23, 2003 10:30pm | #8

   I usually stop at 220 with the ROS wet it. let it dry and sand again with 220 by hand.  This is for anything stained, dyed and/or top coated.  If I'm doing an oil finish I'll sometimes rub the 2nd coat of oil with 320 or higher.  As if this isn't complicated enough, beside grit size you have to worry about the sharpness of the abrasive.  Dull abrasives will burnish the surface.  I try to use raking light and watch for any signs of polishing or burnishing.  If the wood is starting to get shiny, it's time for a fresh disk.  John O'Connell - JKO Handcrafted Woodworking

   Life is tough.  It's tougher if you're stupid - John Wayne

2. 
   Kevin | Jun 23, 2003 10:49pm | #9

   I think the difference is, as Jeff points out in his book Great Wood Finishes and I pointed out earlier, the difference between pigmented stains and dye stains. With pigmented stains the final sanding grit makes a huge difference in how the wood takes the color because the grit choice has a great deal to do with how many places the pigments will have to lodge in and avoid being wiped off. With dye stains final sanding grit choice makes a great deal less difference in how the wood takes the stain. Gel stains would be somewhat of an exception to this because they are designed to have the color sorta float on top of the wood - which is why they work well for blotch-prone woods and also why they tend to not be as transparent as regular wiping stains.

   There are a number of variables that affect the final color. Sanding is one of the most important ones. But, wiping technique can be just as, or even more, important in certain situations... particularly when working with pigmented stains.

   That said... I absolutely do agree with your suggestion to play around with different grits. It's a fantastic suggestion that should shorten the learning curve for less experienced finishers.

   Regards,

   Kevin

3. 
   HowardAcheson | Jun 23, 2003 11:10pm | #10

   I, of course, always recommed that sample boards be made up for any critical finishing job. The point of the tests we did dealt more with adhesion rather than coloring. My own experience, and most of the finishers I know, is that sanding beyond 220 serves little purpose and can be detrimental. For certain woods and stains, there certainly could be exceptions. That's what sample boards will tell you.

   As to gel stains, they are sort of strange animals. For all intents and purposes, they stay on the surface of wood with little or no coloring penetration. They are more like paint or a semi-translucent outdoor stain in that regard. Because of the limited penetration, they work on blotch prone woods. But, they tend to obscure the grain depending on how dark they are and they can chip off of soft woods leaving the lighter wood showing through. I'm sitting at a pine desk right now that was finished with a gel stain and varnish. I keep a dye stick in the drawer to touch up the dings.

   For a maple table I just finished, I used first a dye and then a pigment stain. If the maple was sanded to 220, the pigment stain completely wiped off. Therefore, I sanded only to 150 before staining.