1. Always wash a new brush before you use it. I use Dawn Liquid Detergent in the shop, but you can use most any detergent that doesn’t contain bleach. Bleach will dry out your bristles, just like your hair. I squeeze the detergent into the palm of my hand and then carefully mash my brush bristles into the soap. Rinse with cool water. This will help remove any loose bristles before you begin to paint. This step will prevent those pesky hairs from ending up on your canvas, chair, or trim. Once you have washed your brush smack it on the palm of your hand or on a tabletop to loosen any stray hairs.
2. When you finish painting for the day try to remove the majority of the paint from the brush onto your work surface or a rag before you start to clean the brush. The less paint in the brush, the easier it will be to clean. Rinse the brush holding it by the handle with the brush end pointing down into the water stream. Do not hold the brush head under the water with the bristles pointing up as this will push the paint down into the brush. Pour some detergent into the palm of your hand and gently mash the bristles into the soap, working them up into the brush. Repeat this step till the soapsuds are not the color of the paint and the water runs clear when you rinse out the soap.
3. Always condition your natural bristle brushes after washing. Once my brush has been washed I squirt a small amount of inexpensive hair conditioner into my palm and work it into the brush just like I do when I condition my own hair. Rinse with cool water and shape the brush.
4. Hang your brushes to dry. Do not dry your brushes upside down, i.e. brush head up and handle down. If there is any residual paint in the brush it will settle down into the bristles close to the ferrule and will eventually cause the bristles to break from the build up. I prefer to hang my brushes over the sink to dry. You can use some wire to create “hooks” and hang the brushes off the faucet, thus allowing any water to drip into the sink( see photo). Once dry, the brushes are returned to the brush board (see photo). For my fine art natural bristle brushes, I wash them as I have described and lay them on a shop towel to dry. Once dry they are returned to their proper containers for storage.
While there are some woodworkers who might cringe at the thought of staining wood, there are good reasons why you might want to do so. Some lighter colored woods, such as poplar, alder, beech, and birch do tend to look somewhat bland, and can benefit from a dash of color. Who hasn’t bought a load of wood, only to find, after milling, a disappointing variation in the color – staining can even out the tone of the wood. When building new furniture or cabinetry to match existing pieces, staining might be the only way to blend the two. And, with the price of some exotic woods, such as ebony, reaching stratospheric levels, staining a wood to mimic the look of another species might not seem like such a bad idea. Or, you might simply want to make a really bold design statement. Whatever the reason, enhancing, and even completely altering, the color of wood is a valid, and indeed time honored woodworking tradition.
Bombe Box by Charles Neil
In a nutshell, a stain is a colorant that’s dissolved in some kind of solvent (aka ‘carrier’). The two most common types of colorants are pigments (which cover the wood), and dyes (which penetrate the wood). Some manufacturers combine both pigments and dyes into a single blended stain. The solvent can be alcohol, a petroleum distillate (such as mineral spirits, kerosene, and naphtha), and even the actual finish. A binder is added to pigment stains to help the colorant attach itself to the wood – typically the binder is a resin (acrylic, vinyl, alkyd and the like).
Pigments are fine insoluble particles of inert chemical compounds that can be natural (like iron oxides) or synthetic. Because pigment particles are heavier than the binder, they settle to the bottom of a can – which is why you have to stir these stains before using them. That big glob at the bottom of the can is a sure indicator that the can contains a pigment stain.
Pigment particles lodge in the pores and scratches on a wood surface, exaggerating grain differences. A pigment stain on open porous woods like oak tends to color the early wood more darkly than the denser latewood. However, on closer grained wood like maple, the stain would be much less pronounced because those pores are so small.
While pigments remain closer to the surface of the wood, dyes are absorbed into the fibrous structure of wood, which means the wood will be much more evenly colored. Dyes can be either natural or synthetic. Like pigments they require a solvent, but they don’t use a binder. This is because dye particles, which are about 1/1000th the size of pigment particles, are fully dissolved in the solvent, while pigment particles are suspended in the solvent. Because dye particles are dispersed uniformly throughout the stain they don’t have to be stirred. Like pigments, dyes can be dissolved in water, alcohol, petroleum distillates like naphtha, and in glycol ethers – in which case they’re generally referred to as non-grain raising (NGR) dyes.
While natural dyes are prone to fading, modern synthetic dyes are quite fade resistant (when used indoors). The nice thing about dyes is that you can apply one color directly on top another color. You can purchase dyes either as powders that you dissolve in a solvent (water, alcohol, oil) or pre-mixed in a solvent.
Sometimes you’ll read about people using various chemicals to color wood – potassium dichromate, ammonium hydroxide, sodium hydroxide, along with a range of other suspicious sounding names. Most of these chemicals are both poisonous and caustic, so you should really avoid them. There are so many stain colors available today that you’ll certainly find the right one for your project without compromising your health.
Staining wood isn’t difficult, just a bit messy sometimes. You can apply most stains with a brush or rag. It’s important to note that different woods take stain differently. In fact, even one species of wood will take a stain differently depending on such factors as how the board was cut, the wood grain, and the relative density of the wood. Woods like pine, poplar, alder, and birch are notorious for blotching. For blotch prone wood I apply a thinned coat of shellac before I apply the stain. I also apply shellac on the end grain, which has a tendency to absorb stain like a sponge and darken much more than the surface.
Proper sanding of both solid wood and plywood is a key to achieving optimal results when staining. Make sure that you do a final sanding by hand, in the direction of the grain. Shine a light across the surface of the wood at a 45º angle – it will help you see any imperfections that need attending to. Also sanding the end grain smoother burnishes the pores and reduces their ability to absorb stain. It’s important to remove excess glue completely, particularly around joints.
It’s a good idea to do a trial run on scrap pieces before you begin staining your project. You can experiment by diluting the stain with its solvent (check the label on the can) or laying on two or three coats of stain (allowing each coat to dry in between, of course). You really want to stir the contents of the can thoroughly before applying the stain. It’s important that all the pigment is dispersed in the liquid as opposed to sitting on the bottom of the can. Generally, if a wood is very hard and dense, you should sand to a coarser grit than if the wood is soft and porous. If you sand maple to 220 grit it will absorb very little stain; better to sand the maple to 150 grit. However with oak I tend to sand to a higher grit, often 320.
For important projects I always buy ‘architectural’ or ‘cabinet’ grade plywood. Typically the face veneer on the good side will be one continuous sheet. Building supply stores typically carry a lower grade of plywood. Often, the face veneer is made up of multiple parallel bands of veneer joined together and running the length of the sheet. The bands may not be matched, and more often than not there are even gaps along the glue lines. When stain is applied, the glue lines will likely absorb stain differently than the veneer.
Ban the Blotches
Some woods, such as pine, cherry, and birch, are blotch prone – they absorb stain unevenly. For these I use a gel stain. These thicker stains don’t penetrate wood grain as much as thinner liquid (oil or water based) stains. A second option is to apply a wood conditioner (aka washcoat) before staining. The wood conditioner will help the wood absorb the stain more evenly. You can buy an off-the-shelf conditioner, or make your own by diluting your finish with the appropriate solvent in a ratio of 9 parts solvent to 1 part finish. Remember to allow the conditioner to thoroughly dry before applying the stain.
Lay it On
You can apply stains with a paint brush, foam brush, by rag (my preferred method), or spray. Liquid stains, particularly water based stains, dry fairly quickly, so on large surfaces you want to maintain a wet edge, to avoid lap marks. If they do occur lay on a second coat of stain after the first one has dried. On any project that will likely be exposed to direct sunlight, use a pigment stain – it’s much more lightfast. For projects that require vibrant, bright colors, or where you need to match an existing stain as closely as possible, use dyes. Any stain will result in a darker color if you leave it on longer, or if you re-apply it after the initial stain coat has dried.
Don’t take my word for it – you really need to try several different stains to find the one that suits your needs. Most are available in half-pint sizes for a few dollars a can. Once you’ve latched onto a stain that you like, experiment with it on the woods that you typically build with. Try different finishes on top of the stain as well.
You wouldn’t expect to cut perfect dovetails the first time you try, so you shouldn’t expect to get perfect staining results without some practice. But you will be surprised just how easy it is once you’ve invested a little time in honing your staining skills.
A key factor in choosing a finish for an enclosed space such as a drawer, cabinet interior, humidor or a small room such as a wine or liquor cellar is residual odor. All types of varnishes and lacquers outgas smelly solvents for many days or weeks depending on the thickness applied, the temperature, and the air movement. If you can’t allow that much time, you need to choose another finish.
The two that will leave the least residual odor are shellac and water-based finish. Both contain solvents that evaporate fairly slowly (alcohol in shellac and glycol ether in water-based finish), but they leave almost no odor.
It’s still best to allow adequate time for the finish to dry completely. A way to test is to press your nose up against the finish and take a whiff. If there’s absolutely no odor, the finish is dry.
“A finish should invite you, it should ask to be rubbed and touched and it should feel like warm butter.”
Of any single thing that has been the definitive selling point for my furniture through the years, it has been the finish.
Proper equipment, skill and environment can produce this, but often, for the average guy, it just doesn’t happen to his satisfaction. The simple solution is to rub the finish, but here again, there are many different means of doing that. Some simply are a ‘Witch’s Brew’ that are not successful. Some are very labor intensive and some are so antiquated they just don’t perform. Times have changed. Let’s understand how to effectively finish the finish.
Let’s cut to the chase, with products like pumice, rottenstone, steel wool, polishing compound and so forth, they work but the issue at least for me, has always been inconsistency in sheen. I personally prefer a crisp, clean sheen, I don’t want scratches or dulled, muted sheens. Meaning, even if it is a matte sheen, I want it clear and crisp. My experience and what I teach is that when you use abrasives – products like Micro Mesh – to achieve that it allows us to sand to a consistent sheen.
Compounds like auto body polishing compound are designed to go to one sheen — gloss — so being able to rub to a different sheen is very difficult. If you rub two strokes here, three there, you have changed the sheen, additionally, without proper pads and so forth, it’s all by hand.
Micro Mesh and related products simply put ‘sand’ just like sanding wood, but the consistent grit affords for a consistent sheen, no matter how much you sand, a 1200 grit scratch is a 1200 grit scratch, no matter what, so the sheen is consistent. Additionally, being able to use a 5” random orbit sander is a huge plus, not only from a labor
standpoint, but we also get the random orbit scratch pattern that is far less discernible by the eye, a much improved method.
How To Do It
The first thing you need is something to rub, meaning a good cured finish. Remember you are sanding. Any rubbing product is an abrasive and therefore it’s removing finish, so you have to have enough build-up of finish to allow rubbing. While an oil finish can be improved if needed with a light rubbing, typically a well applied lacquer, shellac, or waterborne finish, often referred to as a film finish, needs nothing more, but any finish can be dramatically improved.
My standard rule of thumb is if I’m going to rub a finish, I apply an extra coat in the case
of an oil finish, or two extra with waterborne. When I apply what would be the regular final coat, as soon as it is dry enough to lightly touch, I apply another.
Because waterborne doesn’t burn into the previous coat as well as a solvent base, if we rub through one coat into the next, we can get what is called ‘Ghosting” or ‘Witness Marks’ which allows you to see the various coats, so if the extra coat is done “back to back” it allows the two coats to better burn in and form one heavy coat, much less likely to cut through.
One quick note, do not rely on rubbing to correct poor prep. If you’re planning to rub a porous wood like oak, mahogany, ash or similar wood, remember the sanding dust and
‘sludge’ can get trapped in deep grain and can produce a white film and if allowed to dry can be very difficult to remove. Keep your surface clean as you go or grain fill it so you have a level surface. You want to have a level finish before you apply the final ‘rub coat’ this insures you’re going to get a nice, flawless finish.
A well cured finish is critical, most finishes 10 to 14 days, you can rub it quicker but the
‘sludge’ from sanding can adhere much easier.
I prefer to ‘dry’ rub when I can, meaning the finish is well cured, and to dust up when sanded, otherwise I use some mild dish soap and water as a lubricant, but again, clean the surface before it dries, the soapy water helps lubricate and prevent clogging of the paper.
Always finish with the sheen you want to end with, meaning use a gloss for gloss, satin for satin and so forth. Shellac, while it can be controlled with sheen altering additives, is always a gloss, but it can be rubbed to any sheen. The issue on a coarse grained wood
is that it is impossible to rub into the grain so a satin sheen with gloss in deep grains doesn’t look so good.
Begin with the finest paper you can, like a 1200 grit or 1500 grit grade or equivalent for simple dust nibs and so forth, if you need to go to a 600 or 800 for more grain leveling, that’s fine, then move through the grits, you can jump grits, no issue. I can go from a 600 to 1000, then to 2000 and so forth. Using Micro Mesh, 1500 is a matte finish and good for leveling, 3200 is a nice satin, 6000 is a nice semi-gloss and 8000 is a super nice gloss. The grits in between allow for control to customize the sheen to your taste so getting a
few extra grits and experimenting is a good idea.
In the ‘P’ grades, 1000 is a matte, 2000 a satin, 3000 semi-gloss and 4000 gloss. Runs, drips and so forth need to be leveled separately using a coarser grit. Super fine
grits like this will float over and not cut it flat. You can take a coarser grit like 320 and a block and level the run or drip prior to using the finer grits. Remember you’re sanding, so be careful on edges and moldings. A quick rub with the grain will usually suffice.
Without question, most rubbed surfaces will have a cloudy or hazy look, not what we want and while waxing helps it will usually streak and smear some. Often this is a result of solvent still in the finish not allowing the wax to fully harden. As well, vigorous
rubbing can melt the wax and cause it to be streaked, using cold water to final rub the wax can help.
Here is where we can borrow from the auto body industry, where rubbing new finishes is a daily thing. Check out the local auto body supply store for some ‘swirl’ remover or
‘glaze’. They can direct you, it’s basically a cleaner/wax product designed to use over fresh finish and a little goes a long way. Apply a light coat and buff it off and your done, you will have a finish that will be a thing of beauty.
If you have questions, please email me at [email protected]
If you have ever sprayed the inside of a cabinet with a fast-drying finish such as lacquer or catalyzed lacquer, you have surely experienced dry spray settling on the surface and causing it to feel rough. The bounce back and turbulence created by the force of the spray keeps the finish particles in the air so long that the finish on the surface has already set up before all of the particles land. And when they do land, they stick to the surface, but don’t dissolve in.
The problem is reduced with HVLP and especially with turbine HVLP because there is so much less force creating bounce back. But in the confined space of a cabinet there’s still almost always some dry spray and some roughness.
Minimal dry spray is difficult to show in a picture. But serious dry spray settling on a surface is easy to show, as on the top half of the panel in the accompanying picture.
I’ve often seen instructions to avoid spraying directly into corners in order to avoid the dry spray. Instead, spray the two surfaces that join to form the corner. In other words, make the solution one of spraying technique. But I’ve never found this to work well.
If you think about it, the real cause of the problem is the fast drying of the finish. Slower-drying finishes, such as varnish and water-based finish rarely, if ever, have a dry-spray problem. So the better trick for avoiding dry spray on the insides of cabinets is to slow the drying of the lacquer or catalyzed lacquer by adding a retarder. This keeps the surfaces wet longer, so the dry spray has time to settle and dissolve in, leaving a smooth feel.
You have to be careful, of course, not to retard too much or you’ll have difficulty avoiding runs and sags. So be observant.
Commercial brands of paste wax are as good as anything you can make yourself, but you may still want to make your own just for fun or to get a specific color or shine. Here’s how to do it.
Grate the wax, or combination of waxes, into a container. Carnauba is available in flake form, so just put it in the container. Add turpentine, mineral spirits or naphtha in the ratio of about ½ pint of solvent to one pound of wax. Then put the container into a pot of water and heat it over a flame, stirring as necessary.
Never put the wax and solvent directly over a heat source; it can catch fire.
After the wax and solvent have combined and cooled, the consistency should be about that of butter in summertime. If you want it thicker, add more wax and reheat. If you want it thinner, add more solvent and reheat.
You can add rottenstone or oil- or Japan-pigment colorants to the wax while it’s still in liquid form and stir it in. Add enough colorant to make the color uniform. You can also add an oil-soluble dye by dissolving and thinning it first in naphtha or toluene.
The degree of shine or gloss corresponds to the melting temperature of the wax. Though the differences aren’t great, carnauba wax has a higher melting point, and thus a higher gloss, than beeswax, and beeswax has a higher melting point, and thus a higher gloss, than paraffin wax.
Thanks for the great tools and products @Woodcraft @KregToolCompany @KennedyMfg @apollospray #EllenDesignOnHGTV @ChipWade Mar 09
Greg Williams has been involved in the furniture industry since 1971, and has been intimately involved in the larger wood finishing world for more than 35 years. He is retired from the RPM Finishes Group where he formerly served as Senior Instructor and Manager of Education for Mohawk Finishing Products and H. Behlen Brothers. He has taught finishing and touch- up and repair to students around the world, and is currently an independent consultant to individuals, companies, and organizations. He can be contacted at[email protected].
What Went Wrong? Predicting Performance
When I first began finishing, on an old gunstock, I read an article that described a simple, time tested procedure and followed it. It worked fine. The first sprayed finish I did was on an army helmet liner, and I was instructed very specifically how to do it. It worked fine. Later, when I worked in a furniture warehouse, I was further educated in procedure by an experienced finisher. I learned how to do specific jobs, but I had little understanding of “why” it was done that way, just that “this way works”.
When I eventually began selling finishing products I was faced with trying to figure out why some things didn’t work, as my customers would challenge me with scenarios of failure. 123
Here’s one. See if you can figure out what happened.
A custom shop fabricated a large, expensive conference table which was made ready for finish just before the Christmas holiday.
The finisher was in the habit of using a high quality lacquer from one manufacturer we’ll call “A” and a good, but less expensive thinner from “B” with acceptable results. Generally the finish work was completed within three days. The first coat of lacquer was applied on December 24 and the last coat was applied January 2. On June 4, the top of the conference table had to be redone. Minor surface abrasion such as that from coffee cups, pens, and even fingernails were leaving highly visible white lines which didn’t even penetrate the topcoat. More vigorous abrasion removed chips of that final coat.
What went wrong?
Thinner B, while perfectly suited to lacquer B, had much less active solvent than required by lacquer A. It reduced viscosity adequately, sprayed and flowed out well, but had less solvent “bite”, or the ability to “burn-in”, than thinner A.
This finisher’s normal procedure was to apply several thin coats in a short time so that previous coats were somewhat open, not fully dried, when subsequent coats were applied. Consequently he normally achieved good intercoat adhesion due to solvent bonding between coats. When the lacquer was allowed to dry over the holidays the last coat lacked the solvent power to adequately resolvate, or liquefy the previous coat allowing the establishment of a solvent bond.
The transportation, labor, and material required to refinish the tabletop cost the finisher more that he had saved all year by buying the cheaper thinner. It isn’t cost effective for a finisher to try to second-guess the manufacturer by economizing on thinner.
Here’s another costly situation. A finisher in a large retail furniture warehouse has a workspace in the northwest corner of the warehouse , enclosed on two sides by 2×4 framing and drywall, with a dropped ceiling at 10 feet.
Above and around this area is the marginally heated warehouse. The heat is turned down to 50 degrees on Saturday night, and since the warehouse is closed on Sunday and Monday, is not turned up until Tuesday morning. The drywall on two sides of the workspace is unpainted, and uninsulated. The other two walls are of unpainted, unsealed cinderblock. The floor is concrete, from which the sealer has long since worn away.
The finisher was charged with doing a burn-in repair on the top of the expensive dining room table. This table was new, had been damaged in an attempted delivery, and now was to be repaired and returned to the unhappy customer. The original salesperson had “promised” the customer that the table would be redelivered within a week. (without consulting the repairman).
The finisher did, indeed, do a masterful job on the repair, but determined that to satisfy this customer he would have to recoat the entire top. The repair was done on Tuesday morning. The air in the repair area was chilly and damp, as it had rained over the weekend and was still drizzly outside. The finisher did have a small space heater in his area, which made it more comfortable and somewhat dryer.
Anticipating the possibility of blushing, he added a small amount of retarder to his lacquer before spraying the top. Unfortunately, there was a slight blush appearing as the lacquer dried. He allowed that first coat to tack off, then resprayed the top, this time with a mixture of about 30 percent lacquer, and 60 percent thinner and 10 percent retarder. Needless to say, the lacquer, thinner, and retarder were from three different manufacturers.
The blush disappeared, and the lacquer leveled well. As it dried, the finisher noticed that the burn-in was now somewhat raised above the surface, as if it had not been properly leveled. He left the shop for the day, expecting to cut down and rub out the top the next morning, or perhaps the following day (Thursday) in time for a Friday delivery. During the evening, the heat was again reduced in the warehouse, and the relative humidity increased as the temperature dropped.
On Wednesday the lacquer appeared too soft to cut down, and the burn-in was still too high, so the finisher decided to respray the top, this time without adding any retarder, in the hope that the next coat would allow him to “level” the burn-in. While that coat was tacking off, it looked good and the burn-in was not as evident as before, so the finisher put the table away to dry and occupied himself with other projects.
On Thursday morning, he almost panicked and quickly called the supplier of the burn-in sticks to ask for help, and to complain about the defective sticks, as the raised burn-in was the most obvious “defect” in the finish at this point.
What Went Wrong.
The retailer was maintaining horrible conditions for spraying lacquer by turning the heat down so far at night, and by constructing the shop area with 4 unsealed walls, two of drywall and two of cinderblock, over an unsealed cement floor. These surfaces hold moisture for days, keeping the humidity high even when the heat is turned up.
To combat the tendency of the lacquer to blush, the finisher added retarder, a slow evaporating solvent designed to hold the lacquer open longer than normal in order to allow the moisture to escape the drying finish. Because he was using thinner, lacquer and retarder from different manufacturers, the technician was not able to easily determine if the “mix” of those products was the best, or even appropriate, for the task. In this case the lacquer thinner was already of a “slower” or “winter” version, that is, it already contained a higher percentage of slower evaporating solvents designed to balance a lacquer which contained less of those solvents. The lacquer, however, did contain a higher percentage of the slower solvents, and was reputed to be quite “blush resistant” under normal conditions.
In the situation we are describing, the lacquer was cold. The thinner was cold. The air in the compressor was cold. The temperature by itself would slow the release of the solvents from the lacquer, and from the burn-in, which was saturated with slow, cold solvents. Adding retarder in this case, while it did help prevent blushing, only exacerbated an already bad situation.
The resins in the burn-in hold solvents much longer than the lacquer, which is one reason that the suppliers of the burn-in material recommend light coats of lacquer over a burn-in. The several coats of retarded, cold lacquer were then covered with another coat, without retarder, which dried faster, almost locking in the unevaporated solvents and diluents in the previous coats, and in the burn-in. Those trapped solvents kept the burn-in swollen, or “raised” above the surface of the wood, showing a “bump” on the surface.
There is a happy ending to this situation though. The customer was called and the problem explained in terms of bad weather slowing down the finish drying. The upside of the repair, for her, was that she would have more coats of a very durable topcoat than the original, or a replacement table, would have, increasing its value and utility. She agreed to wait for another week. In 3 days the “swelling” of the burn-in was gone, as the additional dry time allowed the solvents to escape both the burn-in and the lacquer. On the 5th day the finisher cut down the top with 400A Wet or dry sandpaper, followed by 600, then 1200, sanding with a block and water soluble lubricant. He then used a buffer and fine polishing compound to bring back the original sheen. 137
Using a different thinner than recommended, even a comparably priced, high quality thinner, can cause the lacquer to behave in a different manner than expected. It could lift or dry slowly, or cause stain to bleed through the sealer. Quality reducers are formulated carefully by the manufacturer to complement the solvents and diluents already in that company’s lacquer, sometimes even a specific formulation of that lacquer.
Research and testing under varying conditions have been performed so manufacturers can reliably predict the performance of their finishing materials. While other thinners may perform satisfactorily under some conditions, especially where a fine finish isn’t a high priority, small variances in application conditions can cause unexpected and expensive results. The slowest solvent in a reducer should have good solvency for the resin system. It shouldn’t be a diluent type solvent that could become trapped under the dry film.
Read the directions. Use the coating manufacturer’s recommended additives. Don’t experiment on the customer’s furniture.
Orange peel is an irregular or bumpy surface in a sprayed finish caused by spraying a liquid that is too viscous, or thick, with too little air pressure, or by moving the spray gun too fast or holding it too far from the surface.
When defined in this manner, the methods for avoiding orange peel are obvious.
To reduce or eliminate orange peel, thin the finish or increase air pressure (or both). If you’re using a turbine rather than compressor, you can’t increase air pressure. So thin the finish.
Also, arrange lights so you can see how the spray wets the surface in a reflection as you’re spraying. Adjust your speed and distance so you are spraying an evenly wet coat.
Here’s a trick for avoiding runs when brushing a finish onto a vertical surface. Brush back over the areas where the finish is so thick that it runs to pick up some of the excess with the brush’s bristles. Then wipe off this excess onto a clean cloth you hold in your other hand, as shown in the accompanying picture.
You may need to do this several times to remove enough finish so it doesn’t run anymore.
Always remember that you can’t see what is happening on the brushed surface unless you look at it in a reflected light. You may need to arrange some lights, and you will probably have to move your head around to see a good reflection.