Variations in Pressure Measurements?

SL1 · September 10, 2009, 03:16 PM

I was looking through some of my old reloading manuals and rediscovered some data that Hornady published to show the effects of primer selection, case volume, powder moisture, etc. The pressures were in CUP, and I was a little surprised at how much variability there was in the pressures of 5-shot strings using identical loads in the same chamber. One extreme spread was over 6,000 CUP.

It occurs to me that this is due both to the variability in the actual pressure and also to the variability in the copper crushers used to measure the pressures.

Do any of you have similar data on the extreme spreads of pressures measured with electronic systems? I am wondering if it is smaller. Or, are the electronically-measured peak pressures even more variable than the amounts that the copper cylinder get crushed?

SL1

SQUAREKNOT · September 10, 2009, 03:20 PM

Lyman has a good write up on it in their reloading book.

**Unclenick** · September 10, 2009, 04:17 PM

This is a topic that has come up before. Some examinations back in the 90's showed the copper crushers not only fail to represent high pressure well, but if you sent the same reference loads to be run in a half dozen different copper crushers you'd get different readings. Sometimes even the same copper crusher with a different technician running it would get different readings (I expect that has to do with some folks having a better natural touch for taking accurate micrometer readings). In any event, knowing those crushers used calibrated slugs and still can't get an accurate reading, makes the idea of getting pressure readings by measuring expansion of something as variable and uncalibrated as case head thicknesses is about like measuring a distance by have random strangers come up and offer their eyeball estimates.

In general, since absolute pressures above around 30,000 are higher than their corresponding copper crusher units, I would expect the extreme spreads to grow with them in rough proportion. The crushers may have lousy absolute accuracy, but I expect the same crusher with the same lot of calibrated slugs and the same operator would still have relative accuracy that wasn't too far off the beam. I just can't really be sure of that given that most higher pressure cartridges operate in a non-linear region of the slug's crush response, and given that the static calibration the slugs get clearly does not carry over well to their dynamic performance.

Another approach is possible. I can tell you that by playing with QuickLOAD you can usually get a test load developed that matches the software prediction exactly by first matching case capacity and bore cross-section area, then tweaking the powder burn rate. If you next measure the velocity change made by changing a primer or other component, you can adjust the powder charge in the software to get that same new velocity and have a reasonably good estimate of the actual pressure change required to get there. If loads and their resulting velocities were given in the original data, you could retroactively work out the real effects in PSI.

SL1 · September 10, 2009, 08:15 PM

Unclenick,

Thanks for the info, but it is not what I am looking for. I have some copper-crusher data for 5- and 6-shot strings, so I can see the variation there. I am wondering about the variation in similar strings taken with electronic methods.

For what it is worth, I looked at the copper crusher data that I have and concluded the following:

1. Pressure differences look like they have a real component that is reflected in velocity differences.
2. The residual of the pressure differences look like they have a normal magnitude of about 1,000-to-2000 CUP, EXCEPT
3. When the powder has been (artificially) dried, the residual pressure differences seem to increase to about 2,000-to-4,000 CUP.
4. The change in the magnitude of the residuals due to a physical change in the powder suggests that the residuals still have a real component, rather than just being random variations in crusher hardness, measurement error, etc.

I am wondering if the electronic data is similar. In particular, I am wondering what to think of chronograph readings that show large velocity variations. Are they also large pressure variations? And, what kind of pressure variations can be occurring without any indications in the velocity data?

SL1

bustoff · September 10, 2009, 08:36 PM

Quote:

Do any of you have similar data on the extreme spreads of pressures measured with electronic systems? I am wondering if it is smaller. Or, are the electronically-measured peak pressures even more variable than the amounts that the copper cylinder get crushed?

Large extreme spreads are shown in electronic pressure measurements also. The a-square handloader manual has a informative chapter on pressure measurement and detection. One area shows test results with a es of 7500 psi with a certain powder, primer and bullet. When one component was changed, primer brand in this instance, the es was 3200 psi, all other factors were the same. Again this was measured in psi. There is a section on copper and lead crusher pressure detection and it states that they do a "great job on checking the uniformity of ammunition" . So the 6000 cup es for 5 rounds may well be accurate readings and not errors in detection.

SL1 · September 10, 2009, 09:02 PM

Thanks bustoff,

I wish I could figure-out how to get some of my graphs out of Excel as files so that I could post them here. It would help with this discussion.

With the copper crusher data, I can do a power law correlation to see now much of the pressue variation is represented in the velocity variation. Then I can look at the differences between the correlation line and the data points to see now much of the pressure variation seems to NOT be showing in the velocity data. It is not clear whether that residual variation is real pressure differences or just random errors in measurements of the actual values. But, one of the experiments that I have the crusher data for was a comparison of powder charges at their normal moisture level and after drying to the point that the powder lost 2.5% of its weight. The dry powder exhibited not only higher pressure and higher extreme spread, but also higher residudal variation from the correlation line between pressure and velocity. Since the changes in the residuals were apparently due to a physical change in the powder, I suspect that they are also physically real. That suggests to me that pressures can vary by maybe ± 2000 psi without any velocity variation under "normal" condiitons" and maybe ± 4000 psi under other powder conditions, even when the same charges are used with the same primers in the same cases with the same bullets in the same guns.

Actually, that isn't such a bad result.

SL1

September 10, 2009, 03:16 PM	#1
SL1 Senior Member Join Date: November 8, 2007 Posts: 2,001	Variations in Pressure Measurements? I was looking through some of my old reloading manuals and rediscovered some data that Hornady published to show the effects of primer selection, case volume, powder moisture, etc. The pressures were in CUP, and I was a little surprised at how much variability there was in the pressures of 5-shot strings using identical loads in the same chamber. One extreme spread was over 6,000 CUP. It occurs to me that this is due both to the variability in the actual pressure and also to the variability in the copper crushers used to measure the pressures. Do any of you have similar data on the extreme spreads of pressures measured with electronic systems? I am wondering if it is smaller. Or, are the electronically-measured peak pressures even more variable than the amounts that the copper cylinder get crushed? SL1

September 10, 2009, 04:17 PM	#3
Unclenick Staff Join Date: March 4, 2005 Location: Ohio Posts: 21,063	This is a topic that has come up before. Some examinations back in the 90's showed the copper crushers not only fail to represent high pressure well, but if you sent the same reference loads to be run in a half dozen different copper crushers you'd get different readings. Sometimes even the same copper crusher with a different technician running it would get different readings (I expect that has to do with some folks having a better natural touch for taking accurate micrometer readings). In any event, knowing those crushers used calibrated slugs and still can't get an accurate reading, makes the idea of getting pressure readings by measuring expansion of something as variable and uncalibrated as case head thicknesses is about like measuring a distance by have random strangers come up and offer their eyeball estimates. In general, since absolute pressures above around 30,000 are higher than their corresponding copper crusher units, I would expect the extreme spreads to grow with them in rough proportion. The crushers may have lousy absolute accuracy, but I expect the same crusher with the same lot of calibrated slugs and the same operator would still have relative accuracy that wasn't too far off the beam. I just can't really be sure of that given that most higher pressure cartridges operate in a non-linear region of the slug's crush response, and given that the static calibration the slugs get clearly does not carry over well to their dynamic performance. Another approach is possible. I can tell you that by playing with QuickLOAD you can usually get a test load developed that matches the software prediction exactly by first matching case capacity and bore cross-section area, then tweaking the powder burn rate. If you next measure the velocity change made by changing a primer or other component, you can adjust the powder charge in the software to get that same new velocity and have a reasonably good estimate of the actual pressure change required to get there. If loads and their resulting velocities were given in the original data, you could retroactively work out the real effects in PSI. __________________ Gunsite Orange Hat Family Member CMP Certified GSM Master Instructor NRA Certified Rifle Instructor NRA Benefactor Member and Golden Eagle

September 10, 2009, 03:20 PM	#2
SQUAREKNOT Senior Member Join Date: July 16, 2008 Posts: 263	Lyman has a good write up on it in their reloading book.

September 10, 2009, 08:15 PM	#4
SL1 Senior Member Join Date: November 8, 2007 Posts: 2,001	Unclenick, Thanks for the info, but it is not what I am looking for. I have some copper-crusher data for 5- and 6-shot strings, so I can see the variation there. I am wondering about the variation in similar strings taken with electronic methods. For what it is worth, I looked at the copper crusher data that I have and concluded the following: 1. Pressure differences look like they have a real component that is reflected in velocity differences. 2. The residual of the pressure differences look like they have a normal magnitude of about 1,000-to-2000 CUP, EXCEPT 3. When the powder has been (artificially) dried, the residual pressure differences seem to increase to about 2,000-to-4,000 CUP. 4. The change in the magnitude of the residuals due to a physical change in the powder suggests that the residuals still have a real component, rather than just being random variations in crusher hardness, measurement error, etc. I am wondering if the electronic data is similar. In particular, I am wondering what to think of chronograph readings that show large velocity variations. Are they also large pressure variations? And, what kind of pressure variations can be occurring without any indications in the velocity data? SL1

September 10, 2009, 09:02 PM	#6
SL1 Senior Member Join Date: November 8, 2007 Posts: 2,001	Thanks bustoff, I wish I could figure-out how to get some of my graphs out of Excel as files so that I could post them here. It would help with this discussion. With the copper crusher data, I can do a power law correlation to see now much of the pressue variation is represented in the velocity variation. Then I can look at the differences between the correlation line and the data points to see now much of the pressure variation seems to NOT be showing in the velocity data. It is not clear whether that residual variation is real pressure differences or just random errors in measurements of the actual values. But, one of the experiments that I have the crusher data for was a comparison of powder charges at their normal moisture level and after drying to the point that the powder lost 2.5% of its weight. The dry powder exhibited not only higher pressure and higher extreme spread, but also higher residudal variation from the correlation line between pressure and velocity. Since the changes in the residuals were apparently due to a physical change in the powder, I suspect that they are also physically real. That suggests to me that pressures can vary by maybe ± 2000 psi without any velocity variation under "normal" condiitons" and maybe ± 4000 psi under other powder conditions, even when the same charges are used with the same primers in the same cases with the same bullets in the same guns. Actually, that isn't such a bad result. SL1