Custom Mixing for Specified Sheen

Thoughts on mixing up coatings with desired sheen characteristics, on demand. June 17, 2010

I want to mix a 20 sheen conversion varnish with a 40 sheen in order to end up with a 30 sheen. Does anyone know the ratio? I know it's not equal parts.

Forum Responses
(Finishing Forum)
From contributor E:
Never have done this mix. It would seem equal mix would make 30 sheen. If you add the two together, 20 + 40 = 60. Divide by 2 = 30. You're bringing up 20 by 10, lowering 40 by 10. Come up with 30. If this is not right, I would like to see the calculation. This formula might come in handy sometime.

From contributor T:
Who is asking for a 30 degree sheen? I did a job for a designer in NYC last year who spec'd a 32 degree sheen! I mixed 1.5 parts of a 35 and a shy 1 part of a 15 to approximate. Looked good. Of course she flat out rejected it (demanding client, etc.). So I told her I'd redo a sample, shot out the straight 35, and she loved it. Go figure. My most popular sheens for clients are MLC Dull and Satins 50/50. Make sure you stir the crap out of it before intermix.

From contributor C:
Do yourself a favor and buy a gallon of flattener from your supplier and a sheen mixing chart or a printout from their lab. Start with gloss and let down according to their specs. Remember even the coatings lab has a fudge factor of +/- 2-3% and without a 60 degree gloss meter, I defy any designer to tell the small difference.

From contributor R:
Why not ask the manufacturer to mix it up for you? You should have the manufacturer attach a label to the container just in case the know-it-all designer says "that's not a 30 sheen."

The label will clearly indicate a custom formulation performed at the factory. The last thing you want is to end up making eleventeen sheen samples because the designer thinks the 30 sheen is off just a teeny weeny bit. Once a designer understands that they are responsible for any and all costs associated with the job, they tend to make up their mind rather quickly. Placing the responsibility of a 30 sheen in the hands of the manufacturer leaves you more time to finish the work already at hand.

From contributor C:
That's definitely a better idea, especially if you don't have scales to accurately weigh the ingredients.

From contributor R:
I've adjusted the sheen of regular nitrocellulose lacquer using a tablespoon dipped into the flattening paste. Not the most scientific way to do it, but for a small job it worked out okay. In a large factory environment do they mix different sheens by weight or by the number of scoops added?

From contributor M:
I use Sherman Williams and they will mix any sheen I ask for. Like others posted, if it says "32 sheen" on the can, and was mixed by a large company with strict QC, the designer/architect will not have a leg to stand on. If they don't like the result of the sample, tell them they have to buy the next batch. When you have a pain in the butt client that is demanding many details, make sure your contract is up to date and you cover your butt.

From contributor C:
As far as I know, all manufacturers use the weight standard, usually .001 gram analytical scales or even sometimes .0001 -or 0005. But for sure, for close enough work, any volumetric container will work. I think everyone not in a coatings lab environment would have started out that way - I had a nice set of spoons and cups in the 70's.

From contributor G:
Contributor E's calculation of the questioner's ratio is correct. The calculation for intermixing sheens is the reverse average of parts.

In contributor T's example, 1.5 X 35 = 52.5 + 1 X 15 = 67.5 / 2.5 = 27

For mental arithmetic, I find it easier to work with whole numbers, so the 1.5 becomes 3 and the 1 becomes 2 for a total of 5 parts.

So: 3 X 35 = 105 + 2 X 15 = 30.
105 + 30 = 135 /5 = 27

If contributor T wanted a sheen closer to 32:
5 parts 35 = 175 + 1 part 15 = 190 / 6 = 31.66

From contributor C:
Well, this opens up a nitpicky can of worms. Even the manufacturers give leeway to their formulas because the flatteners themselves do not have precise repeatability within the batch formulation. It is close for sure, but with the use of an 80/60/20 tri gloss meter, the sheen will vary by as much as 3 degrees.

Since most people, even within the coatings community, can't visibly detect a variance of less than 5 degrees by eye, it becomes somewhat moot to try to be more exact than is possible with a coatings offering that is already off by 2-3 degrees one way or the other.

Having had tri-gloss meters at work - believe me, don't waste your time trying to be exact. You never know for sure what the fluctuating minimal gloss difference will be. Nor are the gloss meters perfectly accurate. Being electronic, their results can fluctuate a little also.

But do keep in mind when working with samples that to know for sure what sheen you end up with, the coating has to age until it is free from the bulk of the solvents.

Sometimes people spray and wait till the surface is dry and then do their comparison, only to find out the next day that it's flatter than when they first compared. If given 24 or more hours (with some coatings as much as 3 days), you will know just how close your match really is.

From contributor E:
Contributor G, I understand you're getting 32+/-sheen. How did you get the 5 parts in this? 5 parts 35 = 175 + 1 part 15 = 190 / 6 = 31.66

I was following you until this part.

From contributor G:
That's the reverse average bit. I used 5 parts because that is the number of parts of 35 that go together with 1 part of 15 to average to 31.66.

And - I agree with contributor C that extremely few people can tell the difference between 5 of sheen. Even using a glossmeter, you have to take several readings and average those. Fine-tuning your gloss to less than 5 is moot. Nobody will notice. I'm just answering the question about how to do the calculation.

From contributor B:
If you are going to flatten a CV, be sure to let it sit overnight before you compare the sheen.