Help with Flat Grinding

Real good, Mike. I hope you're doing well also. I'll shoot you an email. Are you still using the same addy?

Brett
 
Same old addy as always, Brett, fitzo1 AT ameritech dot net.

I'm hanging in there.
 
fitzo said:
Mathematical law: the intersection of two planes is a straight line. :)
Click to expand...

More basic geometry:

The intersection of a plane going through the axis of a cone and a surface on that cone is a straight line.

Less basic geometry:
Any other plane that intersects that cone, but not going through the axis of the cone creates a conic. A conic is a non linear curve than can not be described as a radius, and can be described with a ro value. (my terminology may be a bit off here).

The repercussions of this are: if you do not turn your blade as you go around the belly (which is really just a shallow cone) you will not be grinding in a way where geometrically you would naturally have a straight line. Your flat platen grinds straight lines, not parabolas. So you'd be fudging it, which is probably just fine, but you'd end up with a slightly convex surface instead of a perfectly flat line. Or to put it another way, that "cone" will be a bit bulgy...
 
Guys, got to run off to work...again. Man, my day job is starting to really interfere with my knife making education!
Sorry I can't respond to each of your posts right now. Be back on-line later.
Thanks again for all your help.
- Mitch
 
I think we're in agreement, Nathan. Probably the easiest way to demonstrate this would be to either surface grind or mill a blade shape using compound sine chuck with exaggerated angles, keeping the edge thickness constant along the straight edge. The thickness of the edge could then be measured around the curve at various points to give us a real-world idea of the differences one could expect to encounter in a more typical blade cross section.

But, then, I don't personally know anyone with those tools. How 'bout you? :D
 
Wow, thinking about it, you've got access to that CNC mill, don't ya? Hmmm... :D
 
fitzo said:
I think we're in agreement, Nathan. Probably the easiest way to demonstrate this would be to either surface grind or mill a blade shape using compound sine chuck with exaggerated angles, keeping the edge thickness constant along the straight edge. The thickness of the edge could then be measured around the curve at various points to give us a real-world idea of the differences one could expect to encounter in a more typical blade cross section.

But, then, I don't personally know anyone with those tools. How 'bout you? :D
Click to expand...


I have the tools. But the easiest way would be to simulate it in a computer, which I have. Here is a completely computer simulated knife.


http://nathan.broadtime.com/cadknife.jpg
cadknife.jpg


You can experiment with your grinds in CAD and see the different results of different techniques. And what you're speculating is true.
 
So a flat grind isn't really flat. On an edge that curves(belly) you have straight LINES running perpendicular to the edge from edge to spine. The "flat grind" is actually made up of an infinite number of these lines perpendicular to the edge. the only way to have a grind made up of ONE plane is to have an edge that is perfectly straight. If the edge curves at all your grind can't be flat.(actually you can do it but your edge thickness will vary along the blade and that's not usually a good thing.)
 
It would be geometrically impossible to have 2 totally flat planes intersect each other and have the intersection be a curve, To prove this just take 2 round pieces of totally flat steel(metal, wood, hard plastic, or what have you) with the same circumference and try to spread them apart say 15 or 22 degrees and you will find they actually only touch for a very short distance. If you shear off equal parts of the circles straight to form a flat bottom and keep this flat straight touching along its length you will find that as the curve begins a gap must open. There is no escaping this. If you think you have a blade with 2 totally flat planes a curve and a sharp tip try placing a precision straight edge on the sides both horizontally, vertically and at a 45 and see what happens.
 
Jason Magruder said:
So a flat grind isn't really flat. On an edge that curves(belly) you have straight LINES running perpendicular to the edge from edge to spine. The "flat grind" is actually made up of an infinite number of these lines perpendicular to the edge. the only way to have a grind made up of ONE plane is to have an edge that is perfectly straight. If the edge curves at all your grind can't be flat.(actually you can do it but your edge thickness will vary along the blade and that's not usually a good thing.)
Click to expand...

Yes! You get it!
 
Troop I may have missed it in an earlier thread, but did you say where you are from? I live in the Texas panhandle and would be glad to show you if you were close enough. I'm not saying my grinds are perfect, but a few months ago everything just clicked while grinding and I could show you to get you a good starting point. I'd try to explain it but I think I would confuse us all.
 
Nathan, if I understand correctly you modeled this discussion already? Could you give me an idea of the actual change in tip thickness versus the straight segment? I realize it will vary depending on profile and thickness, but I am trying to think my way through just how much change there is.
 
Gentlemen,
espc, Fitzo, Nathan, ib2v4u, Jason, Troop,

as a poster said on another thread when I asked about baloney vs the thread's factuallity, " Now you stop that:D".

I know Fitzo is a maker and trained/educated as a scientist, Nathan is another precision-oriented guy as a machinist/maker......etc.

But all this plane intersecting a plane, and sine vs cosine, and degree of curvality vs plane-ality (I made that up). I spent 30 plus years in marketing and marketing analysis.....you giuys are giving me a headache :D

Can you keep it simple so even a marketing-trained/educated type guy who loves to make knives might have a chance of understanding?:D

Seriously, I seem to always learn from a discussion like this thread....THANK-YOU...and I didn't even ask the question, TROOP did!
:thumbup:
 
ib2v4u said:
It would be geometrically impossible to have 2 totally flat planes intersect each other and have the intersection be a curve, To prove this just take 2 round pieces of totally flat steel(metal, wood, hard plastic, or what have you) with the same circumference and try to spread them apart say 15 or 22 degrees and you will find they actually only touch for a very short distance. If you shear off equal parts of the circles straight to form a flat bottom and keep this flat straight touching along its length you will find that as the curve begins a gap must open. There is no escaping this. If you think you have a blade with 2 totally flat planes a curve and a sharp tip try placing a precision straight edge on the sides both horizontally, vertically and at a 45 and see what happens.
Click to expand...

ib2v4u,

The curve I spoke of in my earlier post wasn't intended to be a circle. I agree, it doesn't work with a circle.
Imagine two lines terminating as they move away from you that are parallel to the ground. Now, imagine drawing infinite lines perpendicular to your original lines towards the ground. Using a constant angle, now imagine that these lines intersect. As you move down the original, ever narrowing lines with your new intersecting perpendicular lines, a change happens. The interesection of these lines will move increasing upward, or away from the ground, as you move towards the termination point. If you view these intersections from a side view, they will form a curve.
In retrospect, this post seems much clearer than my first. :)

Brett
 
fitzo said:
Nathan, if I understand correctly you modeled this discussion already? Could you give me an idea of the actual change in tip thickness versus the straight segment? I realize it will vary depending on profile and thickness, but I am trying to think my way through just how much change there is.
Click to expand...


Fitzo,

Well, assuming you skew the plane as Karl suggests in his description, you can make the tip the same thickness as the ricasso. But being a plane, you draw a line between these two points, and the farther away from that line, the thinner or thicker your blade becomes without regard to edge or spine thickness you want.

Here is a simulation of a blade with that plane set fairly optimally.

http://nathan.broadtime.com/EDGE.jpg
EDGE.jpg


As he indicates, you don't force your taper, you let it become a natural element of your plane placement. However, as you can see by forcing everything to lay on a flat plane the belly becomes very thin. In this simulation the tip and ricasso is .030, but the belly becomes only .005". If you make the ricasso and belly .030, the tip becomes .060". So you see, forcing a flat plane doesn't work except for a straight edge. This effect will be more pronounced on a less "pointy" knife.
 
Excellent info, Nathan. Exactly what I was looking for. I had to think this all through in my itty-bitty head, and numbers just weren't part of the program. :o This conclusion was why I posted my comment earlier about plane intersection being a straight line.

Thank you for sharing!:thumbup:
 
Nathan the Machinist said:
Fitzo,

Well, assuming you skew the plane as Karl suggests in his description, you can make the tip the same thickness as the ricasso. But being a plane, you draw a line between these two points, and the farther away from that line, the thinner or thicker your blade becomes without regard to edge or spine thickness you want.

Here is a simulation of a blade with that plane set fairly optimally.

http://nathan.broadtime.com/EDGE.jpg
EDGE.jpg


As he indicates, you don't force your taper, you let it become a natural element of your plane placement. However, as you can see by forcing everything to lay on a flat plane the belly becomes very thin. In this simulation the tip and ricasso is .030, but the belly becomes only .005". If you make the ricasso and belly .030, the tip becomes .060". So you see, forcing a flat plane doesn't work except for a straight edge. This effect will be more pronounced on a less "pointy" knife.
Click to expand...

Nathan,

This is how I grind a lot of my blades (especially kitchen knives), a continuous bevel from pommel to tip. I have run into the issue with getting a thin belly, what I do to get rid of that and maintain a constant edge thickness is to alter the angle of the bevel. I would have a fairly hard time explaining how I do it, but basically the bevel angle becomes steeper towards the tip, meaning you have more mass behind the edge near the back of the blade (where heavier chopping cuts are made) and a progressively steeper bevel angle towards the tip (where piercing or slicing cuts are made). It also serves to retain more mass near the junction between the blade and handle, which I find helps a lot with keeping the balance point between your index finger and thumb. I would not recommend this type of grind for a chopper or any blade where you want balance and mass more towards the tip, but for a dedicated slicer that feels light in the hand it works very well.

Have a good one,
Nathan (the other machinist)
 
NDallyn said:
Nathan,

This is how I grind a lot of my blades (especially kitchen knives), a continuous bevel from pommel to tip.

snip

the bevel angle becomes steeper towards the tip, meaning you have more mass behind the edge near the back of the blade (where heavier chopping cuts are made) and a progressively steeper bevel angle towards the tip (where piercing or slicing cuts are made).

It also serves to retain more mass near the junction between the blade and handle, which I find helps a lot with keeping the balance point between your index finger and thumb.

snip

for a dedicated slicer that feels light in the hand it works very well.

Have a good one,
Nathan (the other machinist)
Click to expand...

Brilliant

(also) Nathan
 
Brett Bennett said:
ib2v4u,

The curve I spoke of in my earlier post wasn't intended to be a circle. I agree, it doesn't work with a circle.
Imagine two lines terminating as they move away from you that are parallel to the ground. Now, imagine drawing infinite lines perpendicular to your original lines towards the ground. Using a constant angle, now imagine that these lines intersect. As you move down the original, ever narrowing lines with your new intersecting perpendicular lines, a change happens. The interesection of these lines will move increasing upward, or away from the ground, as you move towards the termination point. If you view these intersections from a side view, they will form a curve.
In retrospect, this post seems much clearer than my first. :)

Brett
Click to expand...


Don't work with any curve period.
 
You must log in or register to reply here.
Back
Top