Brush Sanding system; presentation and review

[henrik_o]

Since this came up in a thread some weeks ago, and our new NaxoFlex brushes have arrived, I thought I'd post it as a kind of FYI for those that might be interested in looking at it.

Rather than explain how they work, I'll let some pics do the talking;


8098

Above is a 150mm/width core, with the aluminum end cap detached, showing how the flap/brush strips attach to the core. Changing out a set of strips is very easy and fast, it is done in under a minute.


8099

Another pic of this core; the point of the brushes is to support the sanding flaps which makes the unit much more effective than one with just flaps.
Price for this unit was $185 for the core, and $65 for one complete set of strips.


8100

A unit mounted on a shaper w/ feeder. I think the pic speaks for itself so I won't elaborate.


8101

Another unit mounted on a handheld drill. This type of unit could easily be mounted on a spindle and used on the 'bot.


8102

Cores of the same diameter can be piggybacked to any desired length by detaching end caps and inserting adapters, as shown in this picture. The decorative crownpiece behind the cores was sanded using this system, resulting in a total sanding time of 3-5 minutes, including some slight manual touch up, rather than the ~hour it would have taken to manually sand off all the tool marks (0.5mm step over using a 1/8" ballpoint).

Regarding costs, cores seem to be priced at around $20-35/inch depending on width (the wider the cheaper). Strips are priced in the region of $0.5-1/inch (ditto).

To give a brief user review, I am quite pleased with an important caveat that I do not know how long the flaps will last -- they seem very high quality in the grit, but one never knows.

We are still experimenting with rotational speeds and feed speeds, but so far I am extremely statisfied with the results sanding mdf and softwoods. Especially machined mdf sands like a dream, very fast and perfect finish. Hardwoods, less so, but a coarser grit is plausibly demanded for those. Possibly a different grit medium as well, they offer several options.

The units seem well built, the finish is good, everything goes together exactly as it should without snagging or being loose, the aluminum is well machined and the vinyl (?) part of the core very neat; no gradations or residues.

In terms of operation, changing strips is a breeze. This is the main reason we chose to adopt the NaxoFlex system rather than any of the competitors I'm aware of.

Regarding possible options, all cores are built to order, so you have an enormous range of possible specs. Strips can also be custom made, for example using progressively finer grit along a strip, or using short flaps with coarse grit and long flaps with finer grit on the same strip (or alternating) so you can control finish by how much pressure you apply.

Delivery was on schedule, though due to the custom built nature of the system minimum time is three weeks (in Sweden). The company was very pleasant to deal with, they sent out a representative to demo their products extensively, offered to test sand samples, and stated that if we were not satisfied they would accept a return and credit us the cost. A nice touch was that they called when the order left their facility, and after a few days after delivery their sales rep called to ask how things were going.

We have used our starter units to sand flat surfaces, profiles and '3D' blanks. It seems to work really well for all applications, though complicated 3D blanks require manual sanding in the deepest and/or most confined areas. On one single job, sanding time (including coarse sanding, fine sanding and post-primer sanding) on flat surfaces, edges and profiles of 45 extremely large drawer fronts was reduced from the previous ~9 man-hours to ~1 man-hour: this represents a cost reduction of ~$260 just in labor costs. In other words the time/cost reduction on the first job paid for one complete unit, plus pocket change.

Summa, I'm very satisfied this far and have already started to compile our next order, which will be for a lot more strips of different grits as well as some more cores.

For those that have made it this far into the post, you might be wondering where the 'bot comes into play. Yes, this company also offers a system for use on CNC/linefeed machines with strips rotating on the "end" of a core like the above ones. That will be our next step after covering our bases on the normal cores, if that really works it would be very interesting, I think.

Finally, all of the above is specific to the products of one company, NaxoFlex (http://www.naxoflex.com/default_en.asp), but there are several brands offered by different companies, and I have only looked at the European market.

[joe]

Thanks very much Henrik,

I have a flapwheel which is much smaller and narrower. It comes in handy although it too small. This is a real find. I have a couple of 3D columns where I could sure use one of these.

Thanks again.

[beacon14]

Wow, Henrik, that is an eye-opener. Excellent pictures and explanation. Definitely something to consider.

Thanks for taking the time to make that post.

[desertwoodworker]

Henrik, I went to the NaxoFlex website you entered above and can not find any information on the Flapwheel. Do you have a specific link for the core and the strips?

[henrik_o]

Joe,

quote:I have a flapwheel which is much smaller and narrower. It comes in handy although it too small.
Yeah, we've tried a whole range of products: rigid cylinders, inflatable cylinders, non-backed flapwheels (i.e just flaps with no brushes backing them). They're usually tested, used for a job or two, then ends up on a shelf.

quote:This is a real find. I have a couple of 3D columns where I could sure use one of these.
As I said, we're very pleased this far. There is certainly an amount of technique/skill involved when freehanding 3D work, but with just a little use it just feels right. It's hard to explain, but there's a kind of feedback in using this that once you learn the basics you can 'feel' what you are doing, unlike any other system we have tried.

I'm just starting out, but even with this limited experience I'm nuts over how good it is for sanding off toolmarks in 3D ballpoint machined MDF. I was really worried it would be very aggressive on edges, causing loss of definition in delicate carvings, while not doing much for the toolmarks -- this has been the case with the other mechanical tools I've tried. Yes, you absolutely must learn the tool and how it works, but my first real try after some testing was an amazing success. All those pesky toolmarks, the little lines spaced 0.3-0.5mm apart, most of them gone in a couple of minutes (in deep crevices or confined areas, manual sanding is required, of course).

Maybe I just hit a lucky streak, but I'm very optimistic.

[henrik_o]

David,

quote:Wow, Henrik, that is an eye-opener.
It sure is! We had some old brushes from a now defunct manufacturer, all worn out, but I've been thinking about going this route for quite some time now. Initially, I was going to go with Edruflx, but I was turned off by their business model as well as some technical characteristics, and when NaxoFlex showed up I was sold pretty much on the spot. The system makes sense, it's simple but very flexible, good price and they were super to deal with -- I mean, they deliver solutions for many major brands, such as Swedwood (holding company for all/most of IKEA's manufacturing division), but they still cared about me as a really minor customer. A real class act, which I think is important for us small guys.

I had a good feeling about this, but I also know I've had good feelings about machines/products before and it turned sour in the end. People and companies may be great, but if the product doesn't perform, well... what good is it?

So far, the actual performance of the product has been above ok, it's been great, but I will hesitate to go entirely gospel on it until I learn for how long the flaps last.

[henrik_o]

Dave,

I've talked to them on the phone and they will have a product brochure (in English) ready soon.

They do sell a lot to the US, they've been exhibiting at the IWF I think, but as I understood it mostly to large ventures.

They said they were going to have to look at whether they could expedite small orders through a US intermediary or if they can do it directly.

[henrik_o]

Also, I thought I'd add to the opening post some pics displaying results, since that's what matters in the end.

The following pictures exhibit drawer fronts machined on the 'bot (plus a router table rounding of the inside edges), the unsanded piece below and the sanded piece on top.


8103
Above, a light sanding using 180 grit; it effectively removes the sharpness of the router cuts and 'beards', without compromising the clean lines required.


8104
Again, 180 grit, displaying the slight rounding of the sharp profiles but a slight 'fuzz' still apparent on the end of the sanded piece.


8105

Now with 150 grit: the sanded piece is absolutely clean for priming and will go straight to the paint booth, notice how the definition of the profile has not changed appreciably.


8106
This is how it was done: mount in drill press with improvised dust hood, insert and let work for a couple of seconds, turn around, let work for a second, and so on. RPM ~800.

[gene]

Henrik,
What is the center bore size for the shaft that the unit fits on , and when you have to replace the sandpaper do you have to replace the brushes also , are they made together? Thanks for the post and info . it got the little wheels in my head turning . I wonder how hard it would be to build a small conveyor system and have a few different grit wheels so you could put a course first , med second , and a fine last. would be good for doors and drawer fronts

[sam_harbold]

I have been using this head for about 10 years and it works great. The brushes and Sandpaper are two separate item and can be changed independently.
http://www.sand-rite.com/b-12.html

Here is a little larger unit
http://www.sand-rite.com/flexsander_600.html

Here is another option
http://www.supermaxtools.com/products-superbrush-bush-sanders-24-36-49.php

[harryball]

Very very interesting. I can see using this to clean up bat house kit parts. Right now I use a spindle sander, but this could speed things up tremendously over at the drill press.

Thanks Henrik!

/RB

[3imoh]

For those who are interested...Klingspor makes sanding flap sheets for this sort of thing. They are 9x4 sheets with slits on the edges and a hole in the center. They are basically a giant version of this:
http://www.stockroomsupply.com/The_Sanding_Mop.php

I don't think they have them online, so you need to call or get a catalog.

I have a 1" shaft mounted to a 1.5hp motor spinning a large group of them. They don't have the brushes incorporated into them...that would be nice. The motor mounted in a custom enclosed stand with the sanding mop on top, similar to a shaper. Will try to snap some photos of it sometime.

I also built a dual head version that sanded the top and bottom of material at the same time. It used a power feeder to run stuff through. I was trying to build a poor man's version of something like this:

http://www.quickwood.com/viewsubcategory.asp?maincat=Moulding+Sanders&subca t=CD2%2D300 (http://www.quickwood.com/viewsubcategory.asp?maincat=Moulding+Sanders&subca t=CD2%2D300)

It looks like those guys have now have a portable head with the brushes ...that might be the trick for those who wanted to build something up...
http://www.quickwood.com/viewsubcategory.asp?maincat=Handtools&subcat=F3+Fa st (http://www.quickwood.com/viewsubcategory.asp?maincat=Handtools&subcat=F3+Fa st)

-Howie

[henrik_o]

Gene,

quote:What is the center bore size for the shaft that the unit fits on
Whatever you like. The way I understand it, all cores are built to order, and while they obviously have some standard sizes, I think shaft diameter, width and number of v-slots is totally customizable. Core diameter might not be, I'm not sure.

quote:and when you have to replace the sandpaper do you have to replace the brushes also , are they made together?
Yes. You can see this in the first two pictures in the thread. While this might seem like a waste, it does make for a very simple mechanism that is easy to load/unload, and I would add that once the flaps are worn I think it's a good idea to change the brushes anyway -- it hasn't happen on ours, but I find it hard to believe the brushes would not tend to bend permanently due to the friction of operation (at least on this system, using harder brushes might prevent that but would then make really soft sanding/polishing hard to perform).

quote:Thanks for the post and info . it got the little wheels in my head turning . I wonder how hard it would be to build a small conveyor system and have a few different grit wheels so you could put a course first , med second , and a fine last.
Glad to share. Yeah, I've though along those lines too: the beauty of brushes is that the machine wouldn't require much precision (the way widebelt/drum systems do). Setting up $100 worth of v-belts and pulleys to a main ~$200 motor for the brushes and a secondary ~$125 motor to drive rubber cylinders for conveying should not be that hard. There would be some serious work in making the whole structure etc, but since it doesn't require exact precision it's not painstaking.

I guess the cores (3 each 700mm) would cost something like $3000 which is not trivial, but then you'd essentially have a machine that normally costs ~$15000-20000.

We don't have a need for that right now, but it's absolutely doable.

[henrik_o]

Also, the systems linked by Sam and Howie seem interesting. For me, NaxoFlex was a good choice since I really like the simplicity and ease of customization/operation and they're Swedish.

I think it's important to deal with a company that is there for you and backs you up, so to speak. This is not an exact science, so you need to deal with people that know what they're doing and can guide you right. If we were looking at buying just a brush or two, no big deal, but we wanted to do a comprehensive overview of all our sanding and it adds up. NaxoFlex looks like they are just that kind of partner I want, but everyone should weigh their options as they exist for your specific situation.

I think it helps to sit down and evaluate what you are doing now, what you want to achieve and your budget. My goal was to spend no more than $2000 but cut total sanding time for all operations in half. So far, I've spent $650 in a 'first wave' to attack the biggest time thief, which for us has been profile sanding. Experimenting with these brushes I'm confident that when rounding out this system for ~$400 more I will have cut profile/edge sanding to at least 1/5th of the time used previously.

Such a return on investment is remarkable, but it needs to be said that we were horribly inefficient previosuly: basically hand sanding all edges and profiles.

Since sanding large planar surfaces is a big time slot overall, these reductions on profile sanding don't mean we've cut sanding time to 1/5th in total, more like 1/2. Which is what I set out to do.

I believe reducing the time spent on planar surfaces will be much harder to get a similar time reduction on, and certainly not a similar return on investment. Sanding planes is pretty efficient the way we do it now, and one would need a significant investment to really reduce that time drastically.

Still, rational production dictates that we attack the areas where we can achieve the best returns on investment first, leaving lesser ROIs for later. This goes for everyone, from the huge IKEA factory to the small but high end one man shop. David Buchsbaum has written very well on this in his shopbottools.com blog, and I think it's something we should constantly have in the back of our heads. All operations are different on the face of it, but in the end the same basic rules of rational production applies to everyone.

And yeah, like all sane persons I hate manual sanding, so that's no small part of it either


Edi: the specific post by David B I was thinking about is called More Work In a Day (http://www.shopbottools.com/notes_from_the_sandbox.htm), scroll down to the next to bottom-most blog entry.

[gene]

Henrik,
Thanks for the post, 3000 i can stand 15000-20000 is exactly why i dont have one of these .
Amazing how time directly equates to money . remember time saved is money earned.

[quickwood]

8107

This is the QuickWood version of the Brush sanding system, where the abrasives will slide out of the brushes so only the abrasives need replacement. The abrasives are sitting on an aluminum strip that slides into the brush.

All hubs are custom cut to the size you need and the same with the brushes and abrasives. Delivery time 1 day from order. We are located in Atlanta GA so please add shipping to that time.

We sell about 200 miles a year of this product and have it in every major cabinet shop in USA. The prices are very fair as you only have to buy the abrasives each time. On a semi automatic spindle life time on the abrasives are about 150-200 hours. In automatic opperation like our big sanders you see on www.quickwood.com (http://www.quickwood.com) life time is about 160,000 cabinet doors. Yes I did not make a mistake on that number. We have about 700 machine in North America running this product.

We hope to be able to serve you with our expertice direct from Atlanta. Please give us a call and tell us what you need to do and we will tell you if its the right application for the job. Please call Sand-Tech 770-682-8863

[myxpykalix]

Along the lines of these sanding mops has anyone thought of creating a "sanding toolpath" for their work?
I have been thinking about trying to create a toolpath for spiral files on the indexer after being cut because the way i'm doing it now, is not that efficient.
Here is one example i have seen:
http://www.youtube.com/watch?v=VTSqd_bQ25I
you can skip forward to 8:50 in the video to see the sanding toolpath. It seems pretty simple as far as its movements.

[wberminio]

THis is very interesting,Jack
What really caught my eye was the manual toolchanger.

Erminio

[mitch_prest]

is there any reason a spindle just like that one could not be mounted on a bot. The toolchanger is ideal.. perhaps there is a way to get a cost on just the spindle...

mitch

[myxpykalix]

That video is from Legacy woodworkings new model which is $44,000. OUCH! I did like the wizards that come with the unit, the toolchanger, and the sanding toolpaths but that price and their customer service attitude is 180 degrees different than Shopbots. I have commented to them that they need to follow how shopbot takes care of their customers and stop treating the legacy customers like "red headed stepchildren".

Back to topic....sanding mops