ABRASIVES COMPARISON - multiple types, grits compared by 4 testers

[FenceFurniture]

Over the last several weeks I have accumulated a stash of different abrasive brands and grits and these have been sent out to three other testers, as well as myself. There are 25 discs in all to be compared from 80 to 1500 grit. The four sets didn't quite match up 100% as there was a shortfall on a few discs, and my set was the one the copped most of those shortfalls. This was because I am quite familiar with the missing discs (hence why they are missing ).

A few things to observe before I get to my results:

• The testers were all required to have a Festool ETS 150 sander because the discs that I procured were in the Festool 17 hole pattern, and I wasn't prepared to spend my life's savings on acquiring all manner of DE hole patterns to get some variety happening. This should have no reflection on the results (either an abrasive is good at cutting, or it's not). All of the discs to be tested are 150mm.
• Two of the other testers are professionals, and another is a gifted amateur with a scientific background.
• I have my sander hooked up to a 36mm hose which results in much better extraction than the standard 27mm hose. I mention this because none of the abrasives so far have left any detectable dust on the job.
  
• I am very familiar with two of the brands tested, Jöst and Festool Rubin 2, having used both of them extensively for the past 2+ years. In that time I have developed a preference for the Jöst discs with one exception, the 60g. I have found this particular grit to be not as aggressive as I would have thought, and needs replacing more often than I'd like.
• In particular I have found the finer grits of Jöst (400 and up) to be ridiculously durable. It seems that the finer they get the longer they last (up to 3000). The Festool Rubin 2 range finishes at 220g, so from there up to 500 I have been using the Granat range and have been somewhat underwhelmed by them. I have never used the Festool Platin very fine grits as they are just too expensive by comparison (local RRP is $4.60 per disc compared to the Jöst equivalent "SG2" at around $2.00 including freight from Germany).
• The other discs included are two different types of Klingspor, and Abrasives Industrial, a Perth based company that imports jumbo abrasive rolls from China and processes them here into the various shapes and hole patterns. There were only two types of Klingspor included in the 80g range.
• Also, in the 80g range only there were two types of Jöst - the standard Superpad (yellow) and the newer Abrafilm (and I was omitted from both of these due to short supply).
• I will not supply the various pricings just at the moment until the rest of the results are in.

Some notes and observations on the individual brands and types (and sequenced in the order that I tested them, for want of a better reason):

JÖST

• These abrasives are made in Germany and have to be specifically imported (i.e they are not sold in Australia).
• They have a patented DE hole pattern that suits any sander's holes. They achieve this by covering the disc with hundreds of 2mm holes rather than 9 to 17 larger holes. This means that the holes don't necessarily line up very numerously with the sander's holes. It also results in some dust staying on the back of the disc, and this is no big deal - it does not affect the general extraction at all and in my extensive use of them I have noted that the amount of dust left on the job is the same as the Festool 17 hole pattern - nil that I can wipe off with my fingers.
• The range of grits is from 40 and 60 in Red, 80 to 600 in Yellow, and these are all known as Superpad. The range then goes on from 800 to 4000 in "SG2" and these are on a 3mm thick sponge substrate with no DE holes at all, as the dust is fine enough to be sucked through the sponge at these grits. The same applies to the Festool Platin discs.
• Very occasionally I get some little tears just at the edges of the discs, depending upon what I am sanding. This has never been a problem for me, and doesn't seem to affect performance or leave undue scratch marks.
• The second Jöst range is Abrafilm, and as the name implies, they are a film substrate rather than paper. This significantly increase the strength of the substrate. In the 40 to (I think) 150 range the abrasive is Aluminium Zirconia, and from there up to 3000 the abrasive is a ceramic. Caveat - I need to check the crossover point, and get a better description of the "ceramic".

ABRASIVES INDUSTRIAL

• This company imports large rolls of abrasives from China and cuts them to size in Perth.
• They have two ranges of abrasives. The range we tested is an Aluminium Oxide grit on film substrate, with a grit range 40 to 1200. The other range is a more expensive ceramic grit on film substrate, grit range 40 to 2000.
• The die that has been used for punching the 17 hole pattern has deformed the coarser grits to the point where they will not stick properly to the sanding pad - the hole edges are raised up by up to 2mm which prevents the rest of the disc making proper contact. Care had to be taken to ensure as much contact as possible, so that the disc didn't fly off. Having said that, I did not experience any problems because of this improperly formed hole (on the only grit tested so far - 80g), but this deformity needs to be addressed. It remains to be seen how it affects the more delicate fine grits (the hole deformation is different).

KLINGSPOR AlOx range

• This range is locally available, as Klingspor have an Australian distributor (or branch). I brought my testing discs in from the USA because I had a source and also a box of other goods to be despatched, so it was convenient at the time.
• Manufactured in Poland.
• Grit range is 60 to 1500and available in packs of 10 or 50.
• Stearate coated
• Very large range of sizes, shapes and hole patterns

FESTOOL RUBIN 2

• Limited to shapes, sizes and hole patterns to suit Festool Sanders
• Grit range 40 to 220 (can then switch to Granat, and then to Platin 2 to go up to 4000)
• Probably manufactured in Finland

KLINGSPOR Aluminia Zirconia range

• This is a heavier duty range, which has a coraser grit range of 36 to 220
• 50 packs only
• Stearate coated
• Not sure if they are available locally yet.
• Discs only, with a good range of hole patterns

Ok, that will do for this post. I will post my results once the others have told me they are finished. As the proponent of this testing cycle I feel obliged to do as thorough a job as possible, and as empirically as I can. However, I do not have the required instruments to do a thoroughly scientific job (or the knowledge ). Nor do I expect that the testers will go to quite the same lengths.

So far, I have sanded 1.8 kilometres, that's right 1800 metres, 1000 strokes of 930mm up and back of 115mm wide hardwood.

I have gotten past the 80 grit discs yet.

[LGS]

Brett, if you're going to do this, do it right.

1. All grits should be compared, the fact that you have knowledge of some is irrelevant to this comparison.
2. What are you comparing and what are you comparing to? You should have a reference set of grits.
3. Have you got a reference method and a reference set of grits.
4. Are you using the same board or set of boards for each tester?
5. I believe you should;
a) exclude yourself from the comparison as you bring Jost into Australia and therefore have a vested interest.
b) have a neutral person evaluate the result for the same reason.
6. What is your means of comparison
7. What units do you use to compare results
8. Have you spoken to the testers at all about your opinion of the grits? If so, you may have inadvertently skewed results already.

Show your scientific mate and see what he says. Without proper rigor in the testing, your comparison is merely "he said, she said"


I've just reread your notes and you have clearly stated a preference for Jost pads, your results are already biased.
In addition, you have selectively chosen grits based on price. You are excluding information that may affect the outcome.

You got it wrong again. I won't comment on this evaluation any more. One, because I have a system of sanding which gives me reproducible consistent results using Festool pads and secondly, because the results of this "comparison" are already biased.


LGS

[Evanism]

Such passion about sandpaper!

I agree with FF. This isnt supposed to be some rigorous scientific test. Its four dudes who use a lot of sandpaper and seeing if they can drive better value, results or duration from a disk.

If they evaluate and find Jost, or Festool, or Bunnings Generic Crap is better (god forbid!) then this is OK with me. Its all subjective anyway... just like asking the guy at the tool store which is the best drill. Of course he is going to point to the 5 he has in stock - knowing you really need a WhizzBang 9000 from his competitor.

There is a small pile of mostly used Festool disks here I really know I should throw out, but all I see is money... $1 a disk... It makes me sad. Its a false economy, but if I get just one more blatt out of one... horray!

FF has made no secret of his preference for Jost. They sound very cool. A fine mesh, so holes are irrelevant (reduce inventory, therefore price), long (or good) duration and a good price. Importing is not a problem. This sounds like a winner.

If he has put the thumb on the scale a bit by not being strictly-scientifically-impartial, again, no problem. He has clearly stated he is not a scientist and the methods may be deemed to be arbitrary. BUT THATS THE POINT. All here have purchased things based upon the arbitrary opinion of those we have deemed to trust (sometimes very expensive things, my Laguna SUV is such an example) . There might even be a few shills here who spruik on behalf of products or companies...this is part of the modern dangers of The Internet... things are Not What They Seem sometimes.

Be reasonable, prudent and a little cynical and it should be good. FF has stated he likes Jost and stated clearly why. I don't think he's going to poison the chalice for the simple reason there are 3 other testers. Im sure they will be just as reasonable and come back with results that might reflect their preferences, experiences and tools. Again, this is the point.

Personally, I would have loved to try Jost. I love the Festool types I use (Brilliant, Rubin and Rubin2) and feel its the best sandpaper I've even used. If it can be "beaten", excellent, I know then where to acquire it.

[BobL]

The good thing about this forum is that unlike some of the "tests" posted on blogs the methodology.opinion/results of tester can be challenged. Testers should try to take this in a positive light and use the ideas presented to improve their testing if they wish.

Subjective testing is perfectly fine as long as it is clearly identified as such and I think that FF has sort of done that when we already know he's and fan of Jost and he says

As the proponent of this testing cycle I feel obliged to do as thorough a job as possible, and as empirically as I can. However, I do not have the required instruments to do a thoroughly scientific job (or the knowledge ). Nor do I expect that the testers will go to quite the same lengths.

One aspect that would be essential for a an objective evaluation is for the products being tested to be be completely de-identified, so logos, labels and any colours or textures should not identify the product. This may not be possible so no matter what else is done these types of test will always be susceptible to bias whether it's intended or not.

All that aside it's not clear to me what is being tested.
Is it, rate of stock removal, longevity of abrasive, consistency of finisih, smoothness of finish, . . . . . . ??????

[FenceFurniture]

All that aside it's not clear to me what is being tested.
Is it, rate of stock removal, longevity of abrasive, consistency of finisih, smoothness of finish, . . . . . . ??????

That will become clearer when I put up my results Bob. I've just finished my report on the 80g jobbies and the headings are:
Dust left on the surface
Clogging of the abrasive face
Dust left on the back of the discs
Sharpness to touch after each cycle of 50 strokes, compared to a new disc of each type
Surface finish left on the timber
# of strokes to remove the pencil line
Material removed from the board – thickness reduction


and some of those sections will have photographs.

The ultimate aim is to determine which of them gives the best bang for buck. If a 50% more expensive abrasive lasts twice as long then it's a better buy. However, for a low volume user if they are only available in quantities of 50 or 100 then they are probably cost prohibitive.

The same sort of logic applies to speed of material removal, but this is far more relevant to a pro where time is money.

There are various aspects of the results that may appeal to people in different ways. One little quirky thing, for example, is that the Jöst discs are much easier and quicker (for me) to attach. This is because one doesn't bother trying to line up as many of the little holes as possible with the DE holes. One just whacks it on, and it works, and will work just the same even if it is a bit eccentric. With the 17 hole discsneed to be pretty much smack on, and I have found that the most accurate way of lining up all those 17 holes is to insert three 6mm dowels into three holes and drop it over them. Easy enough but just an extra little PITA every disc change.

[FenceFurniture]

....the methodology.opinion/results of tester can be challenged. Testers should try to take this in a positive light and use the ideas presented to improve their testing if they wish.

Indeed Bob, indeed. I'm all for getting the best possible/most accurate results and if others can contribute some ideas in a courteous manner, then let's go and it will be taken on board.

However, when someone is just not capable of being courteous when it comes to anything I might have to say (because I exist), and will go to any lengths to score an imaginary point in some imaginary war they are waging, even when, as we have seen, the substance has not yet been revealed, then they can only be seen as deliberately inflammatory for the hell of it. As such, they will never be taken seriously, or have their suggestions heeded.

[BobL]

Thanks for the info FF and good on for giving this a go even if it is subjective.

That will become clearer when I put up my results Bob. I've just finished my report on the 80g jobbies and the headings are:
Dust left on the surface
Clogging of the abrasive face
Dust left on the back of the discs
Sharpness to touch after each cycle of 50 strokes, compared to a new disc of each type
Surface finish left on the timber
# of strokes to remove the pencil line
Material removed from the board – thickness reduction

Of these
- The first and 3rd are too dependent on other variables and machinery being used (In particular the DC system being used) to worry about and they certainly wouldn't be significant to me about whether I choose one abrasive over another.
- the second may be worth knowing but is also bound up with machinery. How do proposed to asses that. I would use weight gain by the disc see my comment on #7. I usually don't worry about clogging unless it compromises stock removal rate or smoothness.
- the 4th one is too subjective and like the second one, does it really matter, if the abrasive is clogged or blunt but still rapidly removing stock in a smooth way?
- #5 is fine but how will you quantify it
- #6 us a really good one provided you can generate the same pressure pencil line across the wood. I'd suggest drilling a hole in a piece of wood and jamming the pencil in the hole and then dragging that across the piece of wood being marked. That way you will get the same downward force each time
- #7 is the one I would be most interested in. I doubt you can measure that evenly enough across a piece of wood with a standard calliper. Even if you had a deep throated calliper you would have to measure many points and then get a true average i.e. a lot of work. The way I would do it is by weight. If a 300 x 15 x 12 mm board of density of 0.5 g/cc is used this would weigh ~270g - using a 300g balance that can measure to 0.01g ($10 on ebay) a true average thickness reduction of 0.01/270*12 or 0.00044mm could be rapidly detected. If nothing else this would save a heap of time.

The sort of info that I would be mostly interested in is'
When new, how quickly does the abrasive remove stock
After X square m or Y minutes use, how quickly does the abrasive reduce stock.
Repeat pervious Z times to show trend over time.
To avoid operator the pressure on the sander needs to be applied in a way that is not directly influenced by the operator. Perhaps a weight could be added to the sander and the operator is only permitted to move it side to side.
I would be interested to know this for hard and soft wood and also on clean and abrasive wood.

Going back to #5, the smoothness and consistency of the finish is something I'd be interested in
The average smoothness can be semi quantitatively assessed using friction tests.
There are standardised tests for this invoking a devices called a "turtle"
Basically it measures the force needed to drag a piece of a "standard material" across the surface being assessed.
Place a block of wood on the surface and connect that to a length of string and through a pulley fixed to the end of the surface being assessed.
Then on the end of the string add weights until the block of wood starts to move.
It's a tad more complicated than this but at least its quantitative and largely removes the experimenter from the picture.

I hope you can get some useful ideas from this.

[FenceFurniture]

Yeah I agree that some of those headings a pretty irrelevant (and I concur with your thoughts actually, on which are and which aren't), but they are the observations that I made throughout the process. If they were left out then there would be criticism from somewhere. Two years ago when some Jöst papers were sent out, one tester seemed to think that dust on the back of the disc was a problem and was somewhat critical of them for this, but didn't specify why it mattered.

Some, such as sharpness are very subjective.

As far as I'm concerned there are really only two parameters that really matter - material removal and durability verses cost. One can expect that DE would be the same across the 17 hole papers (why would it be different), and that the surface finish from the same grits would be pretty much the same. However, there are some variations in that second one.

Hey, I'd love to have, or have access to, the proper instruments that can measure these things with accuracy and relevance, but I don't, so I can only give it my best shot. If it was a report for a magazine then it would have to be done with those sort of instruments. All those variables that you correctly speak of can be pretty much ironed out with averaging, which is one of the reasons why I did them all in cycles of 50, repeat, repeat, repeat.

I believe that once the results and opinions of the four testers are up, then people will be able to make up their own minds, particularly if there is more concurrence than not. If my results are consistently at odds with others (bearing in mind that there are another 5 grits for me to test) then I'd say that a claim of bias could be substantiated.

[FenceFurniture]

Btw, the most obvious way to measure material removal is to weight the board after each cycle (think you might have said that), but the problem is having a set of scales that will weigh to a tenth of a gram (which is probably the level required), but also be able to weigh a ten kilo board or whatever the one I used is. I have both types of scales, but not as one unit. My little gemmologist's scales will weigh to a tenth gram, but only up to 100 grams or similar.

So, I'm stuck with measuring thickness.

[Bushmiller]

I can see that any criticism of the test at the moment is premature. Clearly it is an appetite whetner at this stage with the all testers still to submit their results. I can see where laboratory conditions and equipment would be ideal, but reasonably speaking these are not available to us.

So I think we should be content with a general appraisal for the range of abrasives. That in itself is much more than has been commonly available until now. We may have to accept that it is not ideal. I quite like people stating their prejudices. It means they themselves have an awareness they are biased and equally everybody is aware of the same thing and can make allowances if they think it neccessary.

I don't believe Brett's self-confessed bias will prevent him prefering an alternative brand if he finds one to be better. In fact I have a prejudice against at least one of the brands Brett has identified for testing, but it is certainly not my place to mention it at this stage. Indeed, it may be me that is the problem (bad workmen blame their tools etc.etc).

For me, I welcome these evaluations as at worst it provides more information than was previously available and at best it introduces a whole new range of options.

I trust this will be a comparrison of abrasives and not an abrasive comparison. It is up to us.

Holding my breath (in a series of short bursts) for the outcome.

Regards
Paul

[BobL]

I trust this will be a comparrison of abrasives and not an abrasive comparison. It is up to us.
l

10/10 for the joke and the sentiments.

I agree we are not all experts at testing and experiments but there are often very simple tests that can be done to quantify performance.
I think I have said this before but I will say it again. Before folks launch into spending a lot of time testing, it might, and I will stress "might" be useful to run the proposed tests past the forums to get some simple ideas for better quantifying testing.

Some folks on the forum know how to do this naturally, and there are others that have been especially trained and spent many years of their working lives as quantitative testers of some sort or another, should be able suggested ways of quantitative testing that don't necessarily require a "laboratory" or expensive "lab apparatus". We are not usually afraid to ask questions about tools and machines so I can't see why questions can't be asked about ways of testing.

The rate of stock removal I have suggested can be improved significantly simply by using a balance, and the degree of smoothness can be assessed with a friction plate and a few weights.
The degrees of smoothness test using a friction plate is quite time consuming and requires some finessing but is not out of the question for a DIY.
While the results obtained with a friction plate are not as definitive as the stock removal test it still provides a numerical result that can be used for comparison between testers and for punters to make more informed judgements.

[FenceFurniture]

Just got off the phone from Klingspor Australia. Very helpful chap called Paul Hoye. He's going to send me some sample abrasives from the other wood ranges, including some film substrates. These will be in the same grits as the ones to be tested, so depending on how many he sends and the timing I may be able to spin these out to the other three.

So, I have a small conundrum, for the next stage of testing. For me this comes in two parts, as I regard the 80g as a stock remover and it is usually my starting point for "hogging off". Grits above that I just regard as the next grit in getting to the final grit.

That means there are two different sets of values to consider:
1. Pure stock removal (80 grit)
2. Surface condition (all other grits)

and therefore two different test sets. Obviously with both test sets the durability of the disc is critically important (this is all about bang for buck with a good finish ).

So, what I have in mind for the second round of >80grit discs is the following:

• Put the board back through the thicknesser, to get a fresh flat face.
• Before testing each disc, go over the face of the board with enough strokes of the grit underneath the one to be tested to replicate the next stage in sanding. That is to say, before testing the 120g I would go over the board with a 100g disc to leave a 100g scratch pattern and see how long it takes to get it out.
• Put the pencil mark over the top of the 100g (or whatever it is for the grit in question - 320g tests would be preceded by 240g each time etc). My thinking here is that when the pencil mark is removed then that probably coincides with when the scratch pattern from the preceding grit is removed.
• Then see how many times each disc can do the same thing in a satisfactory time to assess durability. That would probably be when the number of strokes required is 25% more than when the disc was new. Maybe a slightly higher % but I can play that by ear and adjust accordingly.

I'll do a bit of a test later to see how well the pencil line removal coincides with when I think I would go up to the next grit in the process.

Of course, when we are sanding a job under non-testing circumstances it is indeed very subjective as to when we proceed to the next grit. We would move up when we think it "feels about right" to. Nothing particularly quantitative about that.

To get a consistent pressure with the pencil I can insert it into a board with a weight on it to maintain the same pressure, and just guide it over the board to be sanded. O'course the pencil will need to be revolved around to keep the line at a reasonably consistent width. The one I have been using is a massive 5mm thick jobbie. The 0.5mm one will scribe to fine a line I believe. I may have a slightly thicker one in the drawer.

So, what say the brains trust?

[FenceFurniture]

ABRASIVES INDUSTRIAL

• The die that has been used for punching the 17 hole pattern has deformed the coarser grits to the point where they will not stick properly to the sanding pad - the hole edges are raised up by up to 2mm which prevents the rest of the disc making proper contact. Care had to be taken to ensure as much contact as possible, so that the disc didn't fly off. Having said that, I did not experience any problems because of this improperly formed hole (on the only grit tested so far - 80g), but this deformity needs to be addressed. It remains to be seen how it affects the more delicate fine grits (the hole deformation is different).

This is what I mean about the holes being raised up:




This is how they should be:

[BobL]

The Pencil method will give some quantitative information which is will be useful provided some way of determining if there is any pencil mark left can be worked out.

Of course, when we are sanding a job under non-testing circumstances it is indeed very subjective as to when we proceed to the next grit. We would move up when we think it "feels about right" to. Nothing particularly quantitative about that.

This is where the friction plate would be really useful because it will measure surface friction The sanding would then be performed until the surface reached a constant friction - further sanding just removes stock and will not improve the so this give the point at which the next grit should be used. The measurements could be a bit tedious.

To be quantitative the testing should be done on the same piece of wood and sanded back to the same surface friction using one brand of paper, otherwise the effect of the previous paper will impact on the result for the next paper. If you want to test the range of grits out as a complete system then that would be a separate test - maybe this is what you are trying to test?

t

[FenceFurniture]

The Pencil method will give some quantitative information which is will be useful provided some way of determining if there is any pencil mark left can be worked out.

Visual will have to do - it's what we would normally do if there was a mark there. If you can't see it but there are traces of it there then it wouldn't matter anyway.

This is where the friction plate would be really useful because it will measure surface friction The sanding would then be performed until the surface reached a constant friction - further sanding just removes stock and will not improve the so this give the point at which the next grit should be used. The measurements could be a bit tedious.

I reckon they would be very tedious. A little earlier I tried the pencil method using a new 120g on a surface sanded previously by a new 100g. It only took four strokes to visually remove it. That means that going through the cycles enough times to find out where the papers are starting to be less useful is going be extremely tedious in itself. I suppose I could speed that up by seeing how long it takes to remove the pencil mark after a cycle of 50 strokes (as there is a reasonable expectation the you'd get 50 strokes out of any disc worth it's salt, regardless of how fine it is).

To be quantitative the testing should be done on the same piece of wood and sanded back to the same surface friction using one brand of paper, otherwise the effect of the previous paper will impact on the result for the next paper. If you want to test the range of grits out as a complete system then that would be a separate test - maybe this is what you are trying to test?

That's what I said I would be doing in the previous post. No, not trying to find the best suite of grits, just the best performer in each grit. However, I think one could reasonably expect that this may well be the same brand over the different grits in the second range of testing. I mean, it would hardly be a surprise to find that the same brand is consistently the best performer.