From: Ron Richardson Sun 2/27/2000
Fellow scratch builders;
A while back I had solicited feedback on foam core cutting. I was having issues with overburn and generally getting the airfoil profile to match to the template. After much trial and error, I have come full circle. I wish to share with those about to make the leap into core cutting. Those of you that have been here may get a little nostalgic at this point.
The reason my cores were coming out thin was due to the kerf (or overburn) that the hot wire made as it went through the foam. Eric Sanders discusses this in the help file with Compufoil. However, I didn’t appreciate the materiality until I saw the results for myself.
The overburn occurs in two ways. First is the part you can see at the edge of the core. It looks just like the kerf a saw makes when cutting through wood. The second is less visible. This is where the kerf gets wider at some point inward from the templates. I had read of this and because the culprit was stated to be templates, and the cure was to space the templates away from the foam (which I did), I did not “dress” the bottom cut before making the top cut.
The error from not taking the time to “dress” the bottom cut amounted to .020". Knowing that a .0004" error in profile will change an airfoil’s characteristics, and that additional errors will creep in throughout the building process, I find that starting with a .020" error is unacceptable.
Having relayed the importance of the cutting sequence, I will now list the steps to get an accurate core (the good stuff starts at step 5). I will not go into setting up the bow/wire/power combination at this time (although I have fresh notes available).
Cut the foam blanks to planform shape. Ensure that the ends are parallel if the templates are intended to be used that way (most are).
Rub, then sand off the foam hairs around the perimeter of the blank.
Set up the lower templates making sure they are aligned with the blank (vertically and horizontally). Also, make sure they don’t move. I’m using sheet metal stands pinned to the workbench to hold the templates.
Fix the blank in position between the templates with a trailing edge stop of some kind (I use a straightedge clamped to the workbench) or some double sided tape. Position some weights on top of the blank to ensure it stays flat during the cut (I use about 10 lb..)
Make the bottom cut first. This is important because the overburn occurs evenly above and below the hot wire. When the top cut is made later, the core will have “settled” into the lower overburn thus offsetting the part of the overburn that eats into the top of the final shape of the foam core.
After making the bottom cut, take the blank up and “dress” the cut. Dressing includes rubbing off all the foam hairs. Vacuuming the hairs. Then you must lightly sand with a long (12"+) block. I use a T-bar sander loaded with 150 grit paper. Sanding does two things. First it gets the rest of the hairs loose. Second and most importantly, it removes the barely visible ridge along the template edge of the cut. You will see this in the way the sheen is removed from the cut at each end of the core. I find that loose beads of foam will pull up during sanding and mar the surface even if you’re careful.
Vacuum again.
Return the blank to the cutting bench. Position the upper templates. Reapply the weights. The core should settle down to a point slightly below the cutting level of the lower templates. I haven’t actually witnessed this, but it makes sense. I’ll check it out tomorrow.
Make the cut for the upper portion of the airfoil.
Repeat steps 6 and 7 on the upper surface of the core.
Check each end of the core against additional printouts of the airfoil.
Remember, the ridge along the template edges and the foam hairs amounted to a .020" error in airfoil shape in my case. The steps above will remove them and allow the settling of the core into the lower bed to self correct for the overburn. I was using 20 ga stainless wire on a 30" bow. Your results may vary.
Again, I got much of this from Eric Sanders via Compufoil. I just wanted to add my two cents worth, share with others, and have a document for my records.
Thermals – Ron Richardson
From: Tom Broeski Sun 2/27/2000
People are always talking about wiping off the fuzz after cutting their cores. I use a very fine, high tension nichrome wire. I get zero fuze and no burnout. I also cut the top of my cores first. This lets the panel stay in position while I switch to the bottom templates. Since, the core is still attached, the alignment is very accurate. I get a perfectly feathered razor edge on my cores. I leave a 1/4" of glass on the wings after glassing for a very sharp trailing edge.
Tom
From: Mark Mech Sun 2/27/2000
The reason my cores were coming out thin was due to the kerf (or overburn) that the hot wire made as it went through the foam.
Another problematic issue for CNC type cutters is overburn due to wire velocity. In order for the wire to enter/exit at the same time one end is moving slower in relationship to the other end (or faster the other way around). This translates into wider kerf at the “slow” end.
Over burning the cores should only be an issue when you are trying to cut as fast as you possibly can, I.E. production. For the guy that is just cutting his own cores it shouldn’t be a problem. Just turn the heat down and cut slow. When I am cutting epp, I am cutting as fast as possible because I have a stack of cores that I want to finish. My drop bars have at least 3lbs of lead on them and I have the heat up high enough to produce a fair amount of smoke.(dont try this at home kids!) Styrene doesn’t need nearly as much heat to cut fast. If you are getting a lot of overburn, you should do a combination or all of the steps below:
A. Turn down the heat B. get a heavier weight. C. get better wire (SS or inconel) and make it very tight. D. use the smallest diameter wire that will hold the needed tension. I use .032 on bows over 3ft. and .022 to .024 on bows from 20" to 3ft. I use .014 on my small parallel bar cutig rig. E. dont use a power supply at the top of it’s range, get one with some head room so it wont vary the current as the cut developes.
Mark Mech
From: Ron Richardson Sun 2/27/2000
A follow-up/reply;
I forgot to mention about wiping off the wire after each cut. I have cut cores with and without doing this and can’t be sure if it really helps. Could be because I left a little foam on the blank ahead of the actual LE (which I cut off afterwards) to help stabilize the wire temp.
Regarding Tom’s reply, nearly everyone said to use stainless wire and cut the bottom first. This is what I have done so I have not experienced the result he describes.
Regarding Aerofoam’s comment, I am doing what he describes. The overburn is not a whole lot. However it is additive and between the “cusping” and the “hairs” for two cuts, the error was the .020" I relayed. This was noted on nearly 18-24 cores (ranging from 7" to 10" chords and 24" lengths) before I finally tried the other way.
Lastly on Stan’s reply, I find that the wider kerf on the slow side of the core does not appear material unless the taper ratio is great. High taper ratios like this are usually found on tip panels.
Remember, I repeated this experiment many times and tested each result to the standard of a fresh printout of the airfoil. I have a high degree of confidence in the results and will use the procedure described religiously from now on.
Thank you for your replies. This kind of dialog makes the exchange the great thing it is.
– Ron Richardson
From: Scobie Puchtler Sun 2/27/2000
This translates into wider kerf at the “slow” end. With that said, I’m not real sure but that “gravity-powered” cutters have the same problem.
They do have the same problem. It is especially well illustrated when cutting wings for a delta in which the root may be twice the tip chord or more. Doesn’t matter what’s doing the driving. In this extreme taper situation, it is obvious that most of the pull work is being done at the root templates, where the wire is doing maximum speed within the system. At the tip, the pull cord is still doing its job, but it barely has to work against any resistance, because the wire has such a long time to eat foam that there is virtually no resistance. In fact strange waviness in the foam at the tip can always be expected to some extent because the hot wire spends so much more time in so much less foam that the surface is affected by the constant radiant heat. Its all kind of fascinating once you get to thinking about it.
Lift, Scobie in Seattle
From: Tom Broeski Sun 2/27/2000
The wire I use is from Omega Engineering in Stanford CT 203 359-1660
It is PN: N180-010-50 FT N180/CR20
I assume it is .010. I use about 26 amps and cut at .4 in per second. I don’t really know any of the technical info. I tried the normal wire for the feather cut machine and it was way too thick for me and had shrinkage near the templates. I cut everything by hand now and find it much quicker.
Tom