Sharpening tools should integrate into a logical system. Here is how and why I have revamped my system to use diamond stones as the workhorses.

The tools

The system uses three DMT Dia-Sharp 8″ x 3″ continuous surface diamond stones, a very fine ceramic finishing stone (about 1/2µ), and the Tormek 10″ grinding wheel. The diamond stones are: 45µ (325 mesh), 9µ (1200 mesh), and 3µ (8000 mesh).

At this size, these diamond stones fit in the same holding device as most waterstones and provide enough length to accommodate a honing guide, if desired. I measured the flatness of each using a Starrett straightedge to be better than 0.0005″ (half a thou) over the whole surface.

Diamond stones cut wonderfully fast, require minimal maintenance, and last a long time. Compared with coarse and medium waterstones, they create less mess, and minimize the chance of transferring grit from stone to stone via the blade or a honing guide.

The processes

A tool makes only infrequent trips to the Tormek to reconstruct the primary bevel. The 10″-diameter grinding wheel creates a mild hollow. I usually grind barely out to the edge or just short of it. Next, I clean up the primary bevel with the 9µ/1200 diamond stone by creating small flats on both ends of the hollow, and sometimes use it to start the secondary bevel (e.g. to start a slight camber). Then I use the 3µ/8000 diamond stone to form the secondary bevel, and finally, the 1/2µ ceramic stone with my diamond nagura to bring it to an exquisite edge.

The 45µ/325 diamond stone supplements the Tormek, such as for forming the primary bevel on certain awkward or small tools, or for adjusting a large camber.

For resharpening on the secondary bevel, which is what we are doing most of the time, I use the finest stone I can, based on how dull the edge is. The idea is to strike an efficient balance that avoids spending too much time on the finer stones while minimizing the depth of the initial scratches that later must be removed.

Thus, if just a bit of touch up is needed, the ceramic finishing stone alone may suffice. More often, I’ll start with the 8K diamond, or, if the edge is quite dull, the 1200 diamond, and work down from there.

When working on the secondary bevel, it is important to use a light touch on the diamond stones. I back off any remaining burr or use the “ruler trick” only on the ceramic finishing stone to preserve the polish on the back of the tool.

The key to this system is that diamond stones comprise the literally central position in the process where their speed and responsiveness quickly get the edge ready for the “money” stone – the finest stone.

For the most critical edges, such as a preparing a smoothing plane blade for the finest work, I finish off the edge by stropping it on leather charged with diamond paste down to 1/8µ, perhaps just to make me feel better.

Though these are my usual processes, the great speed of the diamond stones makes it easy to improvise steps to accommodate different situations such as edge nicks, different steels, edge reshaping, etc.

Why these three grits of diamond stones?

I chose these stones (45µ, 9µ, and 3µ) empirically by experimenting with small stones, basing my judgments on sharpening efficiency and edge performance. The key point is that this set of stones works well for me as a system.

Factors in sharpening stone performance are numerous and complex; grit size is only one of them. A diamond stone performs quite differently than a waterstone of the same nominal grit, so using the same numbers for choosing a series of each type of stone would make little sense. Trial and observation work best.

This set of stones yields for me an efficient balance between minimizing the number of stones and minimizing time. I can maintain proper edge geometry, and get very good responsiveness and feel of the steel on the stones at each stage of sharpening. Overall, the increments in grits seem just about right.

Such choices are certainly subject to personal preferences. For example, many woodworkers may prefer the 8K diamond to be the final stone and finish off the edge with some stropping, or prefer to use larger or smaller increments in grit size.

[Addendum 12/7/16: As of 3/1/16, “Nexabond” products are no longer available but the Nexabond 2500M (Medium set time) formulation is sold widely by DAP products as RapidFuse Wood Adhesive. The short and long set Nexabond formulations are no longer available.]

With this glue, we’ve got a high performance, convenient, fast-setting cyanoacrylate specifically formulated for woodworking. Fine, but there are lots of other good glues, so why do we need this one?

The main advantage of Nexabond’s fast set is that a glued assembly can be brought into the next stage of construction much sooner than with PVAs or most other common glues. The clamps come off, freeing both the clamps and shop space, and the assembly can be manipulated. Depending on the projects you build and your shop circumstances, this can significantly change the workflow.

Maybe. In my shop, which is probably like a lot of other low-volume small shops, speedy glue setting is usually not a big practical advantage. For example, the work flow proceeds just fine as I put aside a glued up tabletop and move to other work.

However, there are other times when it would be very helpful to have the glue set right away. In fact, there are more of these occasions than I first thought. As an example, it is nice to glue up a drawer bottom and have it ready to be raised and fit shortly thereafter with the bonus of not having to worry about moisture or swelling at the glue line. Generally, the advantage of fast setting seems to come up when adding components such as a partition, inside frame, or secondary stretcher to a primary structure where the obstruction of clamps would delay further work.

Moreover, there are situations where clamping small parts is awkward or impossible. With Nexabond, small parts can be held in place by hand until sufficient strength has developed. For example, installing corner blocks in a table went a lot easier with Nexabond.

The option of Nexabond has sometimes changed my approach when otherwise it would not have occurred to me. For example, when making a bent lamination form, successively stacking and pattern routing the layers was a breeze with the glue working at my pace. In general, Nexabond is very handy for making jigs.

The main point is that it’s great to have the option of the quick set. I see Nexabond augmenting, not replacing, PVAs and the other glues in my shop.

I’ve also been finding that Nexabond does a good job on end grain, perhaps better than any other glue I’ve tried. It seems to work well by applying a thin sizing coat to the end grain, waiting 1-2 minutes, scraping it smooth, then gluing the joint as usual. Below are some samples showing that the bond exceeded the strength of poplar but not red oak. That’s pretty good for an end grain-to-side grain bond.

All that said, there are major situations where the fast set is a disadvantage for me. When gluing dovetails or a leg-to-apron assembly, I want the extra time after closing the joint to check for square and adjust the clamp angles as needed. Of even more concern is the possibility of a joint seizing when it is partly assembled as I hurry to clamp it home to the shoulder lines.

In some assemblies where placement and trueness are virtually guaranteed without tweaking, this glue might work fine. We’re told it works well in production shop work.

There are also some joints where wood swelling is part of the strategy and water is therefore an advantage, so for these, I will continue to use PVA glue: biscuit joints, dowelling with dowels that have compressed grooves, and, in my opinion, Dominos.

In summary, as a small shop, low volume, custom woodworker, I definitely want Nexabond in my shop, I trust it, and I will be using it a lot, more than I thought at first, though I do not see it replacing PVAs as my primary glue for major assemblies.

[Note to readers: This series on Nexabond glue can be conveniently viewed in a single page by accessing it from the Series Topics page, to which there is a link just below the header image. There you will also find 19 more series on useful woodworking topics.]

[Addendum 12/7/16: As of 3/1/16, “Nexabond” products are no longer available but the Nexabond 2500M (Medium set time) formulation is sold widely by DAP products as RapidFuse Wood Adhesive. The short and long set Nexabond formulations are no longer available.]

The informal shop tests shown in the previous post suggest that Nexabond glue’s bond strength develops slower in cherry than in the other species tested, red oak and poplar. I asked Peter Stevenson, chemist at Sirrus, the maker of Nexabond, about this. I wondered if the wood chemistry varies a lot between species. Here is his answer, quoted here with permission (emphases mine):

“In fact, the wood chemistry does vary drastically in some cases. One of those being cherry, which is much more acidic than poplar, maple, and oak. The acidic properties of the wood act as a secondary stabilizer of sorts and can decrease the polymerization process. Additionally, even within the same wood species you can see some variation in set time relative to early and late wood. While we have observed some variation in set time, we still see adequate bond strength within half an hour for return to service/processing demands. There will, of course, be some exceptions relative to specific scenarios which may arise.”

This leaves me a bit confused theoretically, because I thought red oak is more acidic than cherry. Maybe it depends on the specific acid compounds. I am not a chemist and fortunately don’t have to be one to do woodworking. What matters is what happens in the shop.

Thus, the practical conclusion for woodworkers, in my view, is that when working with any unfamiliar materials – woods, finishes, hardware, and glue – it pays to do a bit of trial-and-observation in the shop. In the case of Nexabond, it makes sense to take a few minutes and some wood scraps to see how quickly bond strength develops in a particular species before committing to a specific time frame for removing the clamps from a set of glue ups.

Of course, if you are able to clamp the work and keep the clamps occupied for at least about a half hour, then it won’t matter. But five minutes may be too soon in some cases, depending on the joint, how you will handle the assembly, and, as we now learn, the wood species.

Gain direct experience with even with the best tools and materials to use them effectively.

Next: some thoughts on practical applications of Nexabond. (The basics were covered in the first installment.)

[Addendum 12/7/16: As of 3/1/16, “Nexabond” products are no longer available but the Nexabond 2500M (Medium set time) formulation is sold widely by DAP products as RapidFuse Wood Adhesive. The short and long set Nexabond formulations are no longer available.]

Does Nexabond cyanoacrylate glue make strong wood glue joints? In short, yes, but let’s take a closer look at that along with other properties of this glue.

Strength

Sirrus, the manufacturer of Nexabond, has graphs on their website that show has shown that white oak bonds made with their glues are as strong or stronger than those made with “water based wood glue.” They use an ATSM test protocol that involves lap shears and is different from the protocol for Titebond III presented on Franklin’s website.

Testing glue in this manner makes practical sense because the glue bonds in most woodworking joints are primarily stressed in shear. Sirrus kindly supplied me with additional data derived from this test protocol that compares Nexabond to Type I and Type II PVA emulsions, the categories of Titebond III and Titebond II, respectively. Again, they are all nearly identical in strength, though Nexabond 2500S measured very slightly, probably insignificantly, less than the others. The strength of all of the bonds exceeded that of the wood.

Endurance

After soak-bake cycling, Nexabond bonds held up almost as well as Type I PVA bonds and considerably better than Type II PVA bonds. In a moisture-UV torture test, Nexabond outperformed both of the PVAs. Still, Sirrus cautions that their glue is water resistant, not waterproof, and the website states “for interior applications only.”

Woodworkers would naturally like assurance that our glue bonds will survive for decades and generations but, of course, we cannot directly know that yet. Thus, for any new glue, like finishes, there is a degree of uncertainty that only time can fully eliminate. In the meantime, we have to rely on tests and chemists, which seems reasonable.

Impact resistance

I’ve noticed that little blocks that were glued in place with a general purpose CA, in reconfiguring my tool cabinet, for example, can be easily knocked free, usually almost cleanly. For me, that is enough of a clue to avoid those glues for structural joints in furniture.

Nexabond is different. Data from Sirrus shows that it equals the PVAs in the impact resistance of shear joints. Also, even after 30 days of 82°C (180°F) heat, Nexabond bonds held up better than PVA bonds, though I don’t know the practical significance of that.

Cold creep

Glue line flexibility (cold creep), to a degree, is good or bad depending on the application. The flexibility of PVA glue helps mortise and tenon joints survive their inherent cross grain conflicts, while the rigidity of urea-formaldehyde glue keeps bent laminations stable.

Sirrus tells me that, as of my inquiry three months ago, they have not specifically evaluated cold creep performance. They point out that CAs in general are rigid but additives in the Nexabond formula create some flexibility, so qualitatively, performance in this regard should land between PVA and UF glues.

Glue line and clamping

I wondered if the glue line with Nexabond can ever be too thin or if I should be concerned about overdoing it with clamp pressure, such as in an edge-to-edge glue joint where an invisible glue line is desirable. None of their work, they say, suggests issues with too much or too little clamp pressure. Use enough pressure to close the joint and make the parts meet tightly without gaps.

Informal shop tests

Now, here are a few unscientific trials from the Heartwood Joint Destruction Laboratory (no affiliation with NASA). Please don’t interpret these as anything more than me just wanting to “see for myself” in my own shop. I wanted to get a sense of the development of the bond strength in different woods.

All of the joints were carefully fit, glued with Nexabond 2500M, and clamped for 5-7 minutes. I broke apart each joint by placing the assembly in a vise with the glue line just above the jaws and then grasping the top piece with large pliers just above the glue line – basically stressing the joint in tension to failure. I used this method simply because it’s easy to do and tough on the joints.

I broke the first set right after removing the clamps. Shown below, even after setting only several minutes, the poplar joint held perfectly, as evidenced by total wood failure, but the red oak and cherry joints obviously needed more time.

I broke another set one hour after removing the clamps, which is still not enough time to fully cure. Shown below, the poplar and oak joints held perfectly but the cherry joint was not ready.

I let two more cherry joints cure overnight before destruction. One shows about 90% wood failure and the other shows 100% wood failure, in other words, that joint held perfectly.

Again, this is obviously not a rigorous test but Nexabond seems to cure slower on cherry. I repeated the 5-7 minute and one hour breaks with cherry, again making sure the joints met perfectly, and the results were essentially the same as above. Perhaps Sirrus can advise on this.

In any case, I like to do these simple shop trials with unfamiliar glue just as I would with an unfamiliar finish-wood combination.

All in all, I’m very impressed with Nexabond glue. I’m convinced it is quite different – better – than any of the several other CAs I’ve tried over the years. 

As a reminder, this review is unsolicted and uncompensated.

Coming up: let’s consider options for practical applications of this glue in the small woodshop.

I’ve also invited Sirrus to comment on these posts.

[Addendum 12/7/16: As of 3/1/16, “Nexabond” products are no longer available but the Nexabond 2500M (Medium set time) formulation is sold widely by DAP products as RapidFuse Wood Adhesive. The short and long set Nexabond formulations are no longer available.]

Nexabond is a unique cyanoacrylate glue specifically formulated  for bonding wood. Like other CAs, its main advantage is speed. It reaches most of its final strength very quickly, and that can change your shop workflow.

I have never trusted other CAs as mainstream woodworking glues, mostly because of their poor shock resistance. More than relying on data, I could see this myself by easily whacking apart small, simple test joints. I just couldn’t trust those CAs, and trust is everything for a woodworker when it comes to glue. But I’m becoming convinced that Nexabond is different.

First, let’s cover the basics of working with Nexabond. By the way, this review is unsolicited and uncompensated.

The full cure time to maximum strength for all of the Nexabond glues is 4 hours but Nexabond 2500M builds 2/3 of the maximum in only 5 minutes. 2500S is faster, 2500L is slower. They bond wood to wood and many other materials such as glass, plastic, and metal to wood. The manufacturer says it also works on MDF and oily exotics such as cocobolo and teak.

The basic procedure is easy. Apply a small amount of Nexabond – much less than you would for PVA glue – to only one surface and spread. The viscosity of 2500M is a little less than Titebond III, yet it is easy to work with and actually runs less than TBIII because there is simply less glue on the wood.

The joint surfaces must meet snugly because this glue is not a gap filler. I’ve found that there is a comfortable amount of open time so there is no rush to get the parts together.

Clamp the joint using customary force to bring the joint surfaces tight together. The instructions on the website say to clamp for a “couple of minutes” but I clamp for five minutes with 2500M, assured that plenty of strength will build in that time. No accelerator is used. Small pieces like corner blocks can just be held into place by hand for less time.

Aim for minimal squeeze out and use a dry cloth to clean up the excess. Nexabond glue does not contain water so there should be no joint swelling issues.

This is a fast glue, which can be helpful, but the flip side is that parts must come together in the correct position right away because there is very little time for repositioning them once they have met.

The glue has a sharp odor and beware, it bonds skin instantly. Unlike most CA glues, the bottle tip doesn’t clog. Shelf life is listed at 12 months at 72°F.

Coming up: A closer look at shear strength, endurance, impact resistance, and cold creep/glue line flexibility. Then a discussion of practical uses for this glue and a note about end grain bonding.

My small shop requires making efficient use of every bit of floor, wall, and air space, including for dust collection. The setup described here has proven convenient and effective.

The Oneida Mini Gorilla is set on the “mobile” stand but since it will not be moved, I took off the wheels and attached wooden blocks under the platform (see photo above) to save a little space and make the assembly more stable. The whole thing is tucked away just to the right of the table saw, with the filter facing mostly away from the interior of the shop.

Using a 5″-4″ reducer, I attached a 4″ flexible plastic-wire hose to intake port of the collector. This type of hose, available from Rockler, Woodcraft, etc., is durable and easy to handle. A 5″ hose seems more awkward to handle and the 4″ certainly handles all of the chip production from my machines, as discussed in the previous post. It also meets the requirement specified by the manufacturer for airflow through the filter.

I simply bring the hose to each machine as needed, which is very quick to do. The default location is attached to the table saw and with some of its length curled under the saw’s motor hood. Each machine – table saw, bandsaw, jointer-planer, router table – and a dust hood accepts a simple press-fit plastic female connector that is attached to the end of the hose. The ports on some of the machines required modifications for the hose connector to fit uniformly on all of them.

The remote control is very handy, almost a must. It works by radio frequency, not infrared, so no optical line of sight is required.

The steel collection drum holds 17 gallons. I empty it into a 45-gallon trash bag, which makes the job easy. Smaller trash bags made the job considerably more difficult. It is fairly tight quarters getting the drum off and back on the machine but not a big deal. I would have bought the super deluxe infinite capacity drum at an untold price but, um, I was told these have been banned because they defy the laws of physics.

The Mini Gorilla can also be ordered from Oneida with a wall mount bracket. This would not have worked out in my shop but you may want to consider it. It allows the use of a 17, 35, or 55-gallon drum, while the mobile stand is limited to the 17-gallon drum.

Now the steel drum is, of course, opaque, and if you let it fill up such that dust gets backed up higher into the system, you will have a very unpleasant time cleaning the pleated HEPA filter. I made this mistake – twice!

A drum level sensor is available from Oneida but my solution was to drill a hole (approximately 1″ in diameter) in the lid and block it with a rubber stopper with a screw eye handle. When my guardian angel taps on my shoulder, I check on the bin by removing the stopper and peering in with a flashlight or just inserting a rolled up piece of paper or dowel as a depth gauge. It works.

The Mini Gorilla motor can be wired as 110V/16A or 230V/8A. I run mine on a 110V/20A circuit with a dedicated outlet. Keep in mind the collector draws a lot of juice and there will always be another machine running with it, so check your electrical capacity and outlets.

The dust hood shown below clamped to an outfeed stand is helpful at various locations, especially for the storm of debris produced by router mortising with an upcut spiral bit. By the way, the shavings under the bench are just fine for a while where they are, without a dust collector.

For many years, dust collection in my shop had been the basic bag-over-bag single-stage collector familiar to most of us small shop woodworkers. A few years ago I upgraded to the Oneida Mini Gorilla, a cyclone collector, and it has been excellent. I’ve received many inquiries about it from woodworkers considering upgrading their dust collection system, so I hope this report will be helpful.

A cyclone collector is more efficient that a single stage collector because it involves two stages of collection based on particle size. Here’s what happens. The impeller fan whips around the intake material in the big sheet metal cone that give these machines their characteristic appearance. The chips and larger dust particles that constitute almost the entire volume of the intake fall into the collection drum below the cone.

Thus, only the finest dust – the most unhealthful – gets passed on to the air filter, the cage-like cylinder sticking out the side of the yellow cone in the photo above. This permits the use of an incredibly efficient sub-micron pleated HEPA filter without it getting continually overburdened and clogged with debris.

What all this means is that my shop is cleaner, and more important, healthier.

The Mini Gorilla easily handles the biggest producers of chips in my shop, the Hammer A3-31 12″ jointer-planer and the DeWalt DW735 planer. The only limitation is the dust gathering efficiency at the source. Fortunately, the A3-31 and DW735 are great in this regard. The Minimax bandsaw is good while the Saw Stop table saw (vintage 2005) is just fair.

Oneida rates the noise produced by this machine at 76dB(A) at 10 feet. As a practical matter, it seems noisier than what I remember of the old bag-on-bag collector but it’s not bad and certainly not a deal breaker. The Mini Gorilla is very space-efficient – notice the motor on top of the cone – so its footprint is smaller than most small single-stage collectors.

It’s not cheap. The base price is now $780, according to Oneida’s website. A mobile stand or wall bracket, steel drum, and shipping push the total over $1 large.

Oneida’s website has details about the Mini Gorilla and the rest of their large line of dust collectors. This review is unsolicited and uncompensated.

Grunt machines like a good dust collector are not glamorous and there is a natural reluctance to invest in them. However, I think they support a sense of freedom in the shop because they take good care of the dirty work and produce a better working environment. That freedom, in turn, produces more creativity, which is what this is all about.

Next: How I set up the Mini Gorilla in my small shop.

If you are working in a shop with a concrete floor, such as in a basement, consider installing a wood floor. It may be easier than you think.

If you are working in the garage, consider coming indoors and using your hard-earned living space for what really matters to you. For example, banish the TV set to an obscure corner somewhere.

A wood floor is easier on your feet and back. It is also much kinder to a dropped tool, especially an edge tool. The wood floor dampens sound so it’s easier on your ears, and it certainly is a whole lot easier on your eyes. You’ll feel better in a shop with a wood floor and enjoy your time there more. Take it from someone who spent too long here:

Consider a floating installation of engineered flooring. The material is basically plywood topped with a thick ply of beautiful hardwood of your choice, pre-finished with a very heavy duty finish. It comes in strips about 5″ to 7 1/2″ wide, which can contain multiple sections across the width, as seen in my floor in the top photo. The planks are attached to each other but not to the floor below, upon which they simply sit. The planks connect to each other with tongue and groove plus glue, or with a super easy “click lock” connection without glue.

Not tough enough for a workshop? I can tell you my experience. Twelve years ago, I installed an engineered floating red oak floor over concrete in my workshop, which is a few feet below grade. It has held up very well, including no cracks and no separation of the planks, with 600-pound machinery rolling over it, big planks of hardwoods dragged over it, sawdust, dropped tools, and so forth.

I do note that the finish of the particular brand that I installed was later assessed to be more susceptible to denting but it is minor, and the floor still looks great. I also find the finish rather slippery, though this has diminished over time, so I would also keep that issue in mind when choosing material. Ask the experts, let them know you are putting the floor in your workshop, and consider the denting and slippery issues.

Installation involves first, a simple moisture test. Then the concrete floor will probably need to be leveled with leveling compound. Next, a heavy poly plastic sheet is laid down and a thin foam sheet goes on top of that. Then the floor is laid down. Finally, molding and thresholds are installed. The biggest issues are strategies to facilitate laying the flooring, such as starting from the correct side of the room, baseboard heating appliances, and so forth.

You can do this; you’re a woodworker for goodness’ sake.

You do, however, need professional advice. Here in eastern Massachusetts, Hosking Hardwood is well-known (you may recall them from their appearances on the “This Old House” PBS TV series), has an informative website, and offers expert advice.