I've got to cut several precision angles on a manual lathe using the compound, these angles call for +/- 2 Min. accuracy, I've never tried to cut any angles this close and was wondering the best way to set them up? The lathes are equipped with DRO's and I have a sine bar too if that can be used. Oh and we don't have a CMM to check these but our customer does!!! These are pump parts going to a nuclear plant so they are very picky about them being correct!!!
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Largest Manufacturing Technology Community on the Web
Stay Connected:
Setting up precision angles on manual lathes?
- Thread starter Cuda
- Start date
- Replies 40
- Views 19,082
Evan
Titanium
- Joined
- Jan 29, 2004
- Location
- Williams Lake, BC, Canada
The angle you are cutting is the hypotenuse of a triangle that has a segment of the lathe centerline as the base. As such, using a unit triangle with a base of 1 calculate the length of the hypotenuse (secant of angle) for the angle in question. Set the compound to slightly less than the correct angle, make it too shallow. Face off a piece of bar for a reference plane. If the angle is to be cut toward the tailstock then make a facing cut on an inside shoulder that faces the chuck.
Wind the top slide back enough so that it can be wound forward the amount of the hypotenuse. Install an indicator on the tool holder. Touch the indicator to the face created and zero the DRO. To meet the required spec this will require an indicator that can show a touch off with .0001 resolution. Back off the tool using the carriage by 1 unit on the DRO. Wind the top slide to traverse the tool toward the face. If it contacts the face before the calculated distance is reached the angle is too shallow, which is what we expect. Steepen the angle until the indicator show a touch at the calculated distance. Given a resolution of .0001" over a distance of 1" this will provide a sensitivity of better than 1 part in 10000 depending on the angle. Done correctly it will put you inside the tolerance band.
Wind the top slide back enough so that it can be wound forward the amount of the hypotenuse. Install an indicator on the tool holder. Touch the indicator to the face created and zero the DRO. To meet the required spec this will require an indicator that can show a touch off with .0001 resolution. Back off the tool using the carriage by 1 unit on the DRO. Wind the top slide to traverse the tool toward the face. If it contacts the face before the calculated distance is reached the angle is too shallow, which is what we expect. Steepen the angle until the indicator show a touch at the calculated distance. Given a resolution of .0001" over a distance of 1" this will provide a sensitivity of better than 1 part in 10000 depending on the angle. Done correctly it will put you inside the tolerance band.
johnoder
Diamond
- Joined
- Jul 16, 2004
- Location
- Houston, TX USA
0° 2' is close to 0.00058" per inch. If your compound is sloppy it will wave around more than that as you crank the crank.
John Oder
John Oder
jim rozen
Diamond
- Joined
- Feb 26, 2004
- Location
- peekskill, NY
Like john says, you probably can't get there from
here, in a reliable fashion. The machine is the
wrong machine.
The best possible way to get close would be to obtain
one of the correct parts with the correct angle,
mount it in the spindle, and indicate with a good dial
indicator that reads down to tenths and set the
taper attachment up to that. A compound probably
is not good enough to meet those tolerances, but I
would attempt to set it the same way, with a known
good part.
All you would have to do is gage the part at one
height, and have the cutting tool ten thousanths
off that height, and the angle will be out of spec.
Jim
here, in a reliable fashion. The machine is the
wrong machine.
The best possible way to get close would be to obtain
one of the correct parts with the correct angle,
mount it in the spindle, and indicate with a good dial
indicator that reads down to tenths and set the
taper attachment up to that. A compound probably
is not good enough to meet those tolerances, but I
would attempt to set it the same way, with a known
good part.
All you would have to do is gage the part at one
height, and have the cutting tool ten thousanths
off that height, and the angle will be out of spec.
Jim
Carl Darnell
Titanium
- Joined
- Jun 13, 2006
- Location
- Taylorsville Ky
Unless you have a ball bearing taper attachment you will have trouble with it holding that tolerance as well. There is considerable drag on most taper attachments and that imparts a load on the carriage.
The only way may be to tighten the gib on the cross slide and compound untill it is a snug drag and use a known good part to set the taper on the compound.
The only way may be to tighten the gib on the cross slide and compound untill it is a snug drag and use a known good part to set the taper on the compound.
What angles are to be cut? Could you offset the tailstock enough and set-up to a good part or sine bar laying on a precision block? Only if it has centers obviously? Remember to (temporary tighten all gibs). Otherwise this is mute. Rough them and farm them out to grind otherwise it won't happen manually cranking handles.
D
D
jim rozen
Diamond
- Joined
- Feb 26, 2004
- Location
- peekskill, NY
Remember if setting the angle to a readout as evan suggests, or if setting
to a standard part, the angle you produce depends implicitly on the height
of the tool. If you cut a different height than you gage, or if you change the
height of the tool during the work, that will give you an incorrect angle.
Jim
to a standard part, the angle you produce depends implicitly on the height
of the tool. If you cut a different height than you gage, or if you change the
height of the tool during the work, that will give you an incorrect angle.
Jim
Evan
Titanium
- Joined
- Jan 29, 2004
- Location
- Williams Lake, BC, Canada
The setup procedure is totally insensitive to height since it is measuring to a plane surface. The tool should be on center though. It isn't as critical as it may seem and will depend on the diameter of the work. The larger the diameter the less it matters. The rule of small angles says that a thou or two deviation from center height will make not make a measureable difference at a scale 10 times finer so ordinary care in tool centering will do.
Even, you are right with (small angles) but we don't actually know what angles the op is wanting to do
, unless I missed it somewhere. It could be a 45* per side for all we know. All I know for sure is he needs +/- 2 Min. accuracy. Witch even than it won't make a huge differance. But just the wavy cut from a compound and surface finish will eat that up for sure.
D
Peter from Holland
Diamond
- Joined
- Jan 15, 2005
- Location
- The Netherlands
Setting up precision angles is done with a sine bar
Make sure you have a surface on the side of your compound parallel with the motion of the compound
I have a sine bar with a hole in it so I can clamp the sine bar and the gauge blocks to the compound with some elastics
Now indicate the sine bar zero
I agree with previous posters this is a grinding operation or perhaps hard turning is an option
Peter
Make sure you have a surface on the side of your compound parallel with the motion of the compound
I have a sine bar with a hole in it so I can clamp the sine bar and the gauge blocks to the compound with some elastics
Now indicate the sine bar zero
I agree with previous posters this is a grinding operation or perhaps hard turning is an option
Peter
jim rozen
Diamond
- Joined
- Feb 26, 2004
- Location
- peekskill, NY
Evan if I understand your procedure correctly the
first touch is indeed on a plane surface. But the
second touch off is indeed on the side of a round
part.
The point at which it touches off (or zeros to a null)
on the side will indeed depend on the height of the
indicator tip, on the side of the cylinder.
Unless I mistake how you proposed to do this.
Jim
first touch is indeed on a plane surface. But the
second touch off is indeed on the side of a round
part.
The point at which it touches off (or zeros to a null)
on the side will indeed depend on the height of the
indicator tip, on the side of the cylinder.
Unless I mistake how you proposed to do this.
Jim
juergenwt
Stainless
- Joined
- Feb 2, 2008
- Location
- Wheaton, IL.
Grinding Job! Inspection grade surface plate!
RC99
Diamond
- Joined
- Mar 26, 2005
I am with the rough on lathe and grind to finish posters..... What surface finish is required???
Evan
Titanium
- Joined
- Jan 29, 2004
- Location
- Williams Lake, BC, Canada
The setup does not involve the part. All we are looking for is to set the compound to the correct angle. The angle will be correct when the compound is wound the exact distance computed as the secant of the angle and touches the reference surface we faced off at the same time. This is easier to do than it sounds since once we have closely approached the correct angle all we do is loosen the compound and turn it to just touch. Bingo. Angle set.
As for the center height vs the work a small deviation above or below center doesn't make a lot of difference unless the part has a very small radius. The rule of small angles shows that for small angles the cosine of the angle barely changes. To put it another way if the tool is slightly high it is still not going to cut "around the curve" of the part by much unless it nears the center line. For instance, if the target diameter is 1 unit and we are .010 high with the tool the the actual cut will be 0.999945 instead of 1 at the start. How much it is off after cutting half way to center will depend on the angle of cut but for 45 degrees will be about .99978 or a difference in slope of 2 parts in 10,000. If the cut starts on a larger diameter and ends on a larger diameter the error will be even smaller.
Thanks for the input. I was in a hurry this morning when I wrote this, but the parts are an open impeller and it's mating suction bell, the impeller blades are cut to a 30 deg. angle +5 min. tol. and it must match a suction bell with a 30 deg. -5 min. tol. the impeller is 18" in dia. Length of cut is about 4 or 5 in. The 2 angles must fit well together to reduce pumping losses. Surface finish is like a 125 max.
jim rozen
Diamond
- Joined
- Feb 26, 2004
- Location
- peekskill, NY
"turn it to just touch..."
How do you determine that it just touches?
What is the indicator contacting when the second touch
off is occurring?
Jim
How do you determine that it just touches?
What is the indicator contacting when the second touch
off is occurring?
Jim
stannp368
Hot Rolled
- Joined
- Jan 24, 2003
- Location
- Philadelphia, PA USA
When making gages w/ precise angles, I would set the lathe compound using a sine bar, cut the angle as close as I can, then set up a Rivett cylindrical grinder w/ a built in sine plate & finish grind w/ no worry. This is the way to go. Nuclear plant or not, I would not be surprised if the close tolerance was unnecessary as many of them are but if that's what is specified, I would use the above procedure.
Evan
Titanium
- Joined
- Jan 29, 2004
- Location
- Williams Lake, BC, Canada
You set the indicator to the faced surface. You back it away by 1 unit. You crank the indicator back toward the faced surface by winding the top slide the calculated distance which is the secant of the angle required. The indicator will touch and give the same reading as when you first touched off if the top slide angle is correct.
WizardOfBoz
Diamond
- Joined
- Sep 30, 2006
- Location
- SE PA, Philly
This is interesting. I was curious as to what a normal clearance spec would be for an open pump impeller. This source gives a value of 0.015-0.020", and states that every 0.002 inches extra decreases pump capacity by 1 percent. Not sure if this is head or volume or some combination, and not sure if this applies to what is actually a pretty big pump.
Fired up excel, and checked the difference between radius for a 30 degree angle, and radius for a 29°55' angle for a 5 inch cut length. It's about 0.010", or about 0.020" diameter. This +0/-0.020" tolerance is in agreement with the above criterion. By itself, it seems not unreasonable to hold a diameter to this tolerance, but the taper may be difficult. But isn't it the case that the functional spec is not the taper, but the fact that the impeller taper matches the shell taper, that the impeller has appropriate clearance at all points? If so, and assuming that the pump case has an accurate and straight ID taper, then measuring the impeller diameter at several specific points along the length of the impeller should suffice (and you can use a 0.001 indicator, you don't need that 0.00005 one!). The taper should fall within spec if you accurately get 3 or 4 diameters along the length of cut.
A question comes to mind: will the interrupted cut of the impeller vanes improve or degrade the quality of the taper created by a taper attachment? Or no effect?
Best,
Jim
