Phaedrus - Posted - 04/27/2024:  09:01:58

Over a decade ago I started experimenting with sand casting spiders. One thing led to another and things got rather interesting... if not complicated. Unfortunately, in my move from Arizona to N Carolina two boxes of office equipment were “lost”, including my PC and two back-up drives, and I lost the notes and data I had collected. In all, I lost about a year's worth of work. Recently, having time on my hands, I decided to try again on a limited scale. What follows documents the spider sand casting process.



I still have the old forms and some earlier castings.



The first thing needed is a spider form. I use a CAD application to design the spider and cut it on a CNC router. Even still, it takes a lot of hand sanding and filing to reach the final form.



In the past I used a product called “Petrobond” sand in the mold, which does a great job of producing an aesthetically “clean” cast but is quite costly. For this exercise I decided to try “greensand”, which is playground sand mixed with clay and a bit of water. A 50 Lb bag of playground sand needs sifting to remove impurities such as rocks and wood. This yields about 30Lb of usable sand.



For clay, as per several videos on YouTube, I used the recommended kitty litter and ground it to dust using a blender.



The sand, clay and water are mixed.



Somehow all my wood flask molds went missing over the years and I had to manufacture a new one. Everything finally done it’s time to load the mold. An eagle eye may notice that I’m using a different spider form than the one I made earlier. I found a box of my old forms and decided to use one of my favorites. The sand is packed around the form and a sprue hole is added.



In the past I purchased different aluminum alloys and I still have some of them stored. I found some 2021 aluminum that I apparently never tried and decided to use it. The foundry is fired up and metal melted. I don’t have pics of the actual pour as it is difficult (and dangerous) to pour molten metal and snap pictures at the same time. What is shown is the spider removed from the flask after the pour. Also not shown is a second pour of a test strip... more on that later. For the Metallurgists out there note that I don’t do any sort of tempering, annealing or aging.



The raw spider needs a lot more work before it’s finished. I should point out that this casting is not as "clean" as it should be. Using greensand instead of Petrobond was a mistake. The cast is badly pitted. I suspect that the sand was too wet, or the sand itself contained impurities that burned on contact with the molten metal. Regardless, I went ahead and used the casting for this demo. Lesson learned...



At this point I want to ensure that the alloy, thus the spider, is worth going forward. I mentioned earlier that along with the spider I also made a test strip. I want to gauge malleability... is the metal too soft or too brittle to be of use. To do this I measure how much force it takes to bend the 5 x 1/2-inch test strip 30 degrees without breaking. Although I lost my data from previous castings, I do recall that the optimal force was between 35 to 50 pounds. In testing this alloy it took 50 pounds of pressure to begin bending and a bit more force to get it to about 20 degrees before it broke. That put this alloy above what I would consider acceptable. But this cast is for demo purposes, so I'll trod on.



A short note on malleability. If an alloy is too soft it absorbs a lot of the string energy that should be driving the cone. That, and it will collapse more under string tension. If the alloy is too brittle it is difficult to machine and bend... making it very risky to put an arch on the legs without snapping them off. If the metal had made it to 30 degrees without breaking it would have been a viable candidate.



After cleaning up the spider with a file and removing the spue the next step is to mill the underside. Cutting off the spue (that big blob of metal from the pour) leaves a stub that must be removed. Note that different casting methods place the spue at different locations or eliminate it altogether. I do it my way.



Next step is to drill the 1/8-inch hole for the tension screw. After milling a slot for the bridge inserts I enlarge the tension screw hole to 3/16 inch and drill down far enough to allow the tension screw head to rest at the bottom of the slot.



After milling the spider the next step is to add a slight arch to the legs. I shoot for the base of the spider to be .20 to .25 inch above plane. This is where things can go south fast. If the alloy is too brittle a leg will break before it bends. It takes a little practice and patience to perfect this operation, the object being not only to get the correct height but have the tip of each leg as close to plane as possible.



The last step is leveling the legs. I paint the tip of each leg with magic marker then hand sand the spider using a 120-grit sanding disc until the absence of color indicates the legs are even. Note that in my first attempt at leveling a spider, long ago, I used my 12" disc sander. Bug mistake. The spider ripped from my hand and shot across the shop like a Ninja shuriken.



So that's it. Or is it?...



Around the time I began messing around casting spiders a friend asked me to check out a buzzing sound that occurred when he hit the high B and D strings. I was able to determine that the buzz was coming from a spider leg. I removed the spider and checked to see if it was level... it was. Taking a chance, I bent the leg slightly down, off plane, and reassembled. No more buzz and the strings seemed to have an increase in tonal quality.



That got me thinking. For example, using D’Addario EJ42 strings on a 25-inch scale with a 3-degree break angle and total string tension of 205 Lbs, the downward force on the bridge calculates to about 11 Lbs. This force is then distributed to eight legs, each leg (supposedly) putting about 1.3 Lbs of force on the cone. Mathematically this is cut and dry... but what about reality?



A way to verify the math was to build a scale that measured the force of each leg under pressure, and another scale that measured total force on the bridge. After several attempts I was able to build a test platform that, with reasonable accuracy, did both.



With that, I tested the spider I just finished. For comparison I also tested an imported spider I had laying around. Using 10 Lbs of force on the bridge I would expect to get 1.25 Lb force on each leg, ideally. The results of each leg, starting at the 12 o'clock position (the leg pointing up the fretboard) and going clockwise;



Leg      mine       import



1          0.6          .31



2         1.23       1.34



3         1.61       1.71



4         1.11       1.35



5         0.9         1.27



6        1.45        0.39



7        1.53        2.34



8        1.55        1.25



These results are not anomalous. In fact, they are quite typical and raise some serious questions. Remember, both these spiders were leveled... which only guarantees that the spider legs will lay flat on the cone rim. But leveling a spider does not “balance” the load on the cone. Take a moment and think what effect that might have on the cone, particularly in extreme cases. Does the cone warp? When there is unequal pressure on the cone does it attenuate or accentuate string frequencies? Does it even matter?



My short answer to many of these questions is “I don’t know”.



Going back to my friend's problem of a buzzing leg. If a “weak” leg had two adjacent “strong” legs, it’s possible that the cone would be depressed enough so that the weak leg in the middle would barely touch (or not touch at all) the cone rim.



Any change in the sound of a reso cannot be attributed to any one thing done in, say, a set-up. The ingredients that make up the sound “system” are so co-dependent that changing even one thing virtually changes the entire system. For example, if I change one string to a heavier gauge there is an increase in the total string tension, which increases the downward force on the bridge, which changes the break angle thus changing the load on the cone which may throw the tension screw out of whack. Granted this is an extreme example, but it illustrates how co-dependent the system is.



Likewise, if I disassemble a reso, balance the spider legs and reassemble I can’t claim that balancing the legs was the sole cause of an improvement, if in fact there was. In messing with components of the entire system I may have inadvertently corrected a previously unknown problem during reassembly. What I do gain is the confidence that I may have eliminated potential problems.

Tom Jr. - Posted - 04/29/2024:  06:38:19

Very nice work and a great write up about it. This is why we are in the golden age of resos. Serious work being done in all areas. Very interesting concept.

daver - Posted - 04/29/2024:  08:47:52

Few non-production builders delve into reso technology. Fewer share their results. Thanks Jason for your post!

Phaedrus - Posted - 04/29/2024:  09:28:23

Tom, Dave... thanks. It's a labor of love.

SamCy - Posted - 04/30/2024:  10:52:34

Given that the stiffness of a beam is proportional to its width, but proportional to the cube of its depth, wondered what you think of the idea of making the bridge legs narrow and deep, like floor joists. Should be able to retain the same stiffness while reducing weight.

Any thoughts about attending Resogat this year?

Phaedrus - Posted - 04/30/2024:  13:38:43

Sam. Email sent.