From: sfaber@ihlpb.att.com
Subject: more 223 Rem loads with pressures
Organization: AT&T

In a previous post I gave some pressure curve results for
.223 Rem loads taken with a strain gauge amplifier in conjunction
with an oscilloscope.  Here are some more results, a few from
the same setup, and more taken with a portable peak strain reading
amplifier.

The strain gauges are small resistive elements on a amide film that
is glued on to the bare metal outside of the gun barrel chamber.
It measures the expansion of the chamber to a part per million level
(called microstrain.)  All measurements were taken on an AR-15 with
a standard FN chrome light barrel with 1/12 twist, and 55gr Win FMJBT.
All velocities are uncorrected at a distance of about 14 ft.

Last time I gave results showing 26-27 gr loads of WW748 powder
giving strains in the low 300 microstrain level and PMC factory
showing 380 microstrain pressures.  The WW748 powder was notably
slower on the pressure curve.  On suggestion I tried AA2230 powder,
27gr and it showed a curve similar to the factory PMC giving 380
microstrain at 3180 fps.  26gr of IMR4895 gave good accuracy and
3025fps at 360 microstrain.
I was wondering if magnum primers would improve the WW748
characteristics, so I tried 28.2gr of WW748 with a WSR primer and a
CCI magnum primer.  The scope traces were almost identical with
just a very small bulge preceding the peak on the magnum primer
curve.  The peak was 320 microstrain and velocities 3016 and 3030 fps
for the magnum primed.

Next, I was able to develop a low cost portable strain amp using the
latest instrumentation amp chips using 2 9V batteries in a small
box.  It is DC coupled with a bridge input circuit so it can measure
static strain as well as peak hold.

I loaded a series of WW748 loads, AA2230 loads, and took some
factory Norinco, Remington, and Win USA ammo along.  See table
of results.  Note temperature was about 28 deg. F (and snowing).
A graph of velocity vs pressure from a least squares
fit with errors noted gives:

WW748:
vel = 2.66(14) * pres(microstrain) + 2198(47)

AA2230:
vel = 1.93(18) * pres(microstrain) + 2408(68)

This would be over a 2900 to 3300 fps range and 280 to 420 microstrain
range.

Visually on the graph, the highest line is the WW748 and below
it the AA2230.  The AA2230 is obviously a faster powder and gives
a higher pressure for a given velocity, but takes less powder
for a given velocity.  The WW748 reaches case capacity at 29 grains,
limiting the practical obtainable velocity to 3175 or so.
The factory ammo grouped between the two lines at the upper end.

The Norinco was the hottest at 3220 fps ave at 406 microstrain.
Next PMC was at 3210 fps and 396 microstrain with overlapping error
bars.  The PMC uses amazingly efficient powder.
Then Win USA at 3175 fps and 382 microstrain, throwing out
the first point.  The Remington was in a class by itself at
2985 fps and 328 microstrain.
The factory ammo gave a .9% velocity std dev where the handloads
were at about .5% average.

I measured case capacities for all the cases mentioned and found they
all gave 28.2-28.4 gr H2O with bullet seated, except the WW cases,
surprizingly, which held 27.7 gr.

I wondered why the Remington was so mild, and then noticed the Speer
reloading book states that military 5.56mm has a different chambering
than commercial 223 Rem. and military ammo may develop excessive
pressures in 223 Rem guns because of this.  Maybe it has something to
do with the freeboring and initial gas leak due to differences in
throat dimensions.  In a previous post I estimated a 1 grain of
powder loss from initial gas leak for my AR loads.
In any case beware that the Chinese Norinco ammo is labled 223 Remington
and not 5.56mm, yet it is at the full military pressure.

Another interesting thing to do is graph log( pressure) vs log (charge)
to get the pressure exponent of the powder.
The WW748 formed a nice line yielding 4.39(7) for the pressure
exponent.
AA2230 was not quite so nice and gave 4.75 +- 1.1

If you use the Lame equations for closed end cylinders to estimate the
pressure that would produce these strain levels on the chamber, you
get 42,500 - 43,600 psi for 400 microstrain.  This seems a bit low.
One could probably use the military load as a calibration point if
you knew what it was supposed to be.  All the military loads seem
to be right around 400 microstrain.




Scope traces:	temp 68 deg F
Type	charge		strain	velocity
----	------		------	--------
PMC	26.6 gr		390 	3233 fps
Rem	26.0		340	3073
SS109	26.6		400	2991

AA2230	27.0		380	3180
AA2230	26.5		330	3058
AA2230	26.0		320	2991
WW748	28.2		330	3103

Peak reading amp:	temp 28 deg F

Type	charge	    number	pressure	velocity
----	-----	    -----	--------	--------
PMC	26.6 gr		4	395(16) 	3210(28) fps
WW USA	27.2		4	383(13)		3175(14)
Norinco	27.2		5	406(7)		3178(16)
Rem	26.0		3	328(4)		2985(20)

AA2230	26.5		5	349(15)		3054(20)
AA2230	27.0		5	387(8)		3149(9)
AA2230	27.5		5	408(14)		3202(17)

WW748	27.5		5	292(6)		2966(9)
WW748	28.2		5	325(14)		3066(18)
WW748	29.0		5	368(7)		3178(16)


Steve F.

From: bartb@hpfcla.fc.hp.com (Bart Bobbitt)
Subject: Re: Ballistics ( books ? )
Organization: Hewlett-Packard Fort Collins Site

David Post (post@hpfcso.fc.hp.com) wrote about my comment on strain gage
pressure measuring by Oheler:

:  >If you're interested in internal ballistics, Oheler has a strain gage
:  >system for measuring your own rifle's pressure.  The gage is epoxied to
:  >the barrel and wires from it go to the system.  A pressure curve is
:  >plotted and stored on a disk.  Your PC then reads the disk and plots
:  >the pressure curve along with other neat and interesting data.

: It must plot relative pressure, right?

It plots actual pressure.  See below.

: I never heard of a strain gage system that did not have to be calibrated,
: even with a prepared surface.

The barrel metal type and thickness at the point where the strain gage
is epoxied on is entered into the system.  That establishes the base for
the actual pressure values to be computed for.

: Of course, I AM getting old....

Well, so is everybody else that's still alive.  Isn't that great?  And
we are all doing it at the same rate.  Some folks just started sooner or
later than others.

BB

From: gmk@falstaff.MAE.CWRU.EDU (Geoff Kotzar)
Subject: Re: Ballistics ( books ? )
Organization: /etc/organization

In article <C405vx.5Kq@fc.hp.com> bartb@hpfcla.fc.hp.com (Bart Bobbitt) writes:
#David Post (post@hpfcso.fc.hp.com) wrote about my comment on strain gage
#pressure measuring by Oheler:
#
#:  >If you're interested in internal ballistics, Oheler has a strain gage
#:  >system for measuring your own rifle's pressure.  The gage is epoxied to
#:  >the barrel and wires from it go to the system.  A pressure curve is
#:  >plotted and stored on a disk.  Your PC then reads the disk and plots
#:  >the pressure curve along with other neat and interesting data.
#
#: It must plot relative pressure, right?
#
#It plots actual pressure.  See below.


I just got off the phone with Ken Oehler and we discussed his Model 43
system. According to him the system plots "relative" or "estimated"
pressure. Pressures are estimated from the strain data by running them
through the Lame' thickwalled pressure vessel formula with an unconstrained
end. They are "relative" because his system is in fact uncalibrated. You
need to know that not every strain gage is applied exactly along its intended
direction. For this pressure estimation system, the gage is supposed to be
oriented circumferentially; if you don't align it correctly, you introduce
an error that can be calibrated out. If you don't calibrate the system then
the error remains. These usually are very small though as a rule because the
misalignments are small.



#: I never heard of a strain gage system that did not have to be calibrated,
#: even with a prepared surface.

This statement is correct, absolutely. Noone would manufacture a transducer,
which is what the pressure barrel becomes, without calibrating it.


#The barrel metal type and thickness at the point where the strain gage
#is epoxied on is entered into the system.  That establishes the base for
#the actual pressure values to be computed for.


These values for the geometric parameters are simply needed for use in the
Lame' formulas. They only allow the user to estimate the pressures. Basically
it is easier to think in terms of "psi" than in terms of "micro-strain" and
that is what the conversion does. It allows you to deal with a unit that has
some intuitive meaning to you. You know that you are not supposed to load
cartridge "A" to more than 52,000 psi but what strain levels are allowed?

The "base for the actual pressure values" is an absolute pressure reference:
a calibration, if you will.

According to Dr. Oehler, you can fire factory rounds in your rifle and accept
whatever estimated pressure you get as your pressure reference and then state
that your loads produce more or less pressure than factory loads and by roughly
how much. But that is all.

I hope this helps clear up some of the misconceptions netters hold about the
capabilities of the pressure measurement system Oehler is marketing.


geoff kotzar

From: sfaber@ihlpb.att.com
Subject: Re: Oehler 43 (was Ballistics ( books ? ))
Organization: AT&T

From article <1ode0oINNc6@usenet.INS.CWRU.Edu>, by
gmk@falstaff.MAE.CWRU.EDU (Geoff Kotzar):

# I just got off the phone with Ken Oehler and we discussed his Model 43
# system. According to him the system plots "relative" or "estimated"
# pressure. Pressures are estimated from the strain data by running them
# through the Lame' thickwalled pressure vessel formula with an unconstrained
# end. They are "relative" because his system is in fact uncalibrated. You
# need to know that not every strain gage is applied exactly along its intended
# direction. For this pressure estimation system, the gage is supposed to be
# oriented circumferentially; if you don't align it correctly, you introduce
# an error that can be calibrated out. If you don't calibrate the system then
# the error remains. These usually are very small though as a rule because the
# misalignments are small.


After obtaining the demo disk for the Model 43, I noticed he
uses a somewhat arbitrary "offset" value of 7000 psi when computing
the pressure with the above formula.  Apparently the Lame' formula
gives a lower pressure than it should, so he adds in an offset.
I wondered if it was really used as an added offset, or was converted
to a scale factor of some kind.  After analyzing the accompanying
raw data I found it really was a simple offset added in.
I have a problem with this, since the strain gage reading should be
proportional to pressure and not involve an offset.  If the Lame'
estimation is off, then it should be a straight scale factor that
should be applied.

The 7000 psi may be close to the observed difference between the
formula in the 50000 psi region.  It seems consistent with my
measurements where the formula gave 43000 psi for a full load that
you would expect to be around 50000 psi, but I claim if you just
add in the offset to the formula value, it will cause errors
with lower pressure measurements.

Steve

From: sfaber@ihlpb.att.com
Subject: Re: Oehler 43 (was Ballistics ( books ? ))
Organization: AT&T

From article <29760@dog.ee.lbl.gov>, by bercov@bevsun.bev.lbl.gov (John
Bercovitz):

# ##After obtaining the demo disk for the Model 43, I noticed he
# ##uses a somwhat arbitrary "offset" value of 7000 psi when computing
# ##the pressure with the above formula.  [stuff deleted]
#
# #Just an intuitive guess: this could be a correction for the stress
# #supported by the case.  I don't have the yield stress and modulus
# #on hand (John Bercowitz, please come to the front...)
#
# I looked in an old post I wrote and sure enough, for mild pressures,
# the case supports 7000 psi of the total pressure.  Nice intuit there,
# Lloyd.  8-)

Yes, excellent, it sure explains things nicely, Thanks!

I had dismissed the idea of the case holding very much pressure
due to an erroneous calculation.  After looking up the formula
in Roark's for thin walled vessels, and using John's value of
63,000 psi from an old posting for the yield stress of the brass,
the results are for .223:  P =  R/t*s  where  R is the radius of the
shell, t is thickness of the wall and s is the stress on the wall.
For t= .015 in,  R=.365/2 in and s= 63000,   P = 5180 psi
which is in the ball park, especially since I believe the yield stress
value was a minimum value.  What do you mean by "for mild pressures",
John?

Steve Faber

From: bercov@bevsun.bev.lbl.gov (John Bercovitz)
Subject: Re: Oehler 43 (was Ballistics ( books ? ))
Organization: Lawrence Berkeley Laboratory, California

In article <1993Mar24.173757.3143@cbnewse.cb.att.com>
sfaber@ihlpb.att.com writes:

##From article <29760@dog.ee.lbl.gov>, by bercov@bevsun.bev.lbl.gov
(John Bercovitz):

## I looked in an old post I wrote and sure enough, for mild pressures,
## the case supports 7000 psi of the total pressure.  Nice intuit there,
## Lloyd.  8-)

#Yes, excellent, it sure explains things nicely, Thanks!
#I had dismissed the idea of the case holding very much pressure
#due to an erroneous calculation.  After looking up the formula
#in Roark's for thin walled vessels, and using John's value of
#63,000 psi from an old posting for the yield stress of the brass,
#the results are for .223:  P =  R/t*s  where  R is the radius of the
#shell, t is thickness of the wall and s is the stress on the wall.
#For t= .015 in,  R=.365/2 in and s= 63000,   P = 5180 psi
#which is in the ball park, especially since I believe the yield stress
#value was a minimum value.  What do you mean by "for mild pressures",
#John?

Well, that was an error.  The cartridge I did the calculation on was a
mild pressure cartridge.  So I threw that bit of pressure information in
gratuitously and definitely in the wrong place.  In other words, I *(^%%$ #$,
Steve.  It makes no difference what the pressure of the propellant is, the
case will make its contribution according to its diameter, wall thickness,
and state of temper.  Thanks for catching this error.

John Bercovitz     (JHBercovitz@lbl.gov)

From: sfaber@ihlpb.att.com
Subject: Re: Strain guage info wanted
Organization: AT&T

From article <C7C4on.5Fp@news.cso.uiuc.edu>, by cbl@uihepa.hep.uiuc.edu
(Chris Luchini):

# Hi all,
#    I need to get as much info as I can on the strain guages that
# are used to find read barrel pressures.
#
# 	o What is the nominal resistance
# 	o What is the variance of the resistance (dR/dP?
# 	o What is the temprature dependence of the resistance.
#
# Thanks

Resistance:  depends on your electronics.  Standard gage resistances
	are 120 ohms and 350 ohms, but you can also get 1000 ohms.

	350 ohms are a good buy and give you a little more voltage
	output with less power consumption than the 120 ohm ones

variance of resistance is  strain * gage factor
	gage factors are around 2

	strain on typical barrels is around 400 to 600 micro in/in

temp dependence:  I think around 1 ppm / deg F,
		gages are compensated for thermal expansion of the
		metal the gage is designed for.

		temp dependence of gage factor is 1.2 % per 100 deg C typ

From: sfaber@intgp1.att.com
Subject: Re: [LONG] [TECHNICAL] .308 Strain Gage Pressure Results

#
#Long article alert! (It didn't start that way..... ;-)

Loved the article.  The extra background  made it much more interesting
and worth the extra length.

#I have not finished analyzing the data, but I will present what I have
#so far, in the hope that it may be of interest.  If anyone wants to
#look over my data, point out errors or oversights, or suggest interesting
#test scenarios, I'd be real interested in hearing.
#
#The 'pressures' listed here should not be treated as CUP, nor should
#they be treated as PSI.  They do not take into account the pressure
#contained by the cartridge case (estimated to be 5-6000 psi), nor
#do they take into account any errors in our "mike and multiply"
#technique of generating strain--->pressure conversion factors.
#

The 5-6000 psi figure was obtained by measuring the hardness of some
cases and using a graph of hardness vs yield stress for cartridge brass
and the case diameter and thickness, calculating the resultant
pressure.   This compares to a value of 7000 psi usually used by
the Oehler system.  This offset added to the pressures below is the
best guess at chamber pressure in psi.


#Jensen Lot 13			5	52769 +- 2000
#
#This is the excellent 1992 Palma Match ammunition, shot at the 1992
#Palma Match in Raton, New Mexico.  It is manufactured by Jensen's
#Custom Ammunition in Arizona.  Jim was Captain of the Canadian Palma
#team, and as a result has a "quantity" (he won't tell me how much)
#of this ammunition.  It is a special lot of Winchester brass (weighing
#166 grains per case), 44.8 grains of IMR4895, a Federal 210 Match
#primer, and the Sierra 155 grain Palma bullet, seated to roughly
#2.79" overall length.  It has a reputation of being a very hot load.
#Jim wanted to make sure that his ammunition developed less pressure
#than Jensen's.  Apparently this ammo has been tested, and was within
#SAAMI spec, albeit near maximums.  The question that I haven't heard
#answered is whether or not this testing took place in a SAAMI test
#barrel, or whether it took place in a chamber/throat/barrel configuration
#likely to be seen in a competitor's Palma rifle.

These loads are good data points for some forthcoming internal
ballistic calculations.

#
#DC Lot 61C 42.7 4895/155	5	45671 +- 2000
#
#DC Lot 63A  41.5 Rel7/ 110 Hdy.5	54339 +- 4000
#
#DC Lot 63C 41.0 4064/180	5	45349 +- 1500
#
#This is another load that uses the SB79 primed case.  I wanted to try IMR
#4064 and some heavier bullets than I would normally shoot out of my
#14" twist rifle, mainly due to Bart's praise for IMR 4064.  This load
#used 180 grain Speer flat base hunting bullets.  Previous testing indicated
#excellent groups at 100 yards (as good as I was getting with the Sierra
#155's), with no indications of bullet tumbling.  I was amazed that these
#bullets would stabilize in my barrel; I'm going to try the much longer
#(i.e., more difficult to stabilize) Sierra 190 BTHP Matchking that is
#so popular under U.S. Palma Rifle rules.
#
#DC Lot 63B 42.7 4064/180	5	50368 +- 1000
#
#This is 1.7 more grains than the above load, resulting in a pressure
#increas of 5000.  Note that IMR4064 seems to give very uniform
#peak pressures.  Perhaps there is something to what Bart says..... ;-)
#

Pretty uniform.  My tests with 2 gages on the same gun show about a 1.5%
std. dev system accuracy level.  Most all of the error is attributable
to the strain gage/glue system and not the electronics.  I used epoxy, and
Daniel used cyanoacrylate.  If they are about the same, then I expect a
standard dev. of about 750 with a 50000 pressure value would be about
the best you could hope to get even with perfect ammo.  You are pretty
much approaching that.

#DC Lot 63D 38.0 H335/180	5	49003 +- 2500
#
#DC Lot 63E 38.0 H335/155	5	34112 +- 2500
#
#
#DC Lot 63F 42.7 H335/180	1	65059
#
#This is a beyond-maximum load!  I fired this with the knowledge that it
#may damage my rifle.  Jim, Dave and I all looked away when I fired this
#shot.  This was defintely, by far, the highest pressure ever shot in my
#rifle.  The bolt required very substantial effort to open (just shy
#of hammering), and the case was stretched .004" more (in headspace)
#than a regular pressure load fired in my chamber.  We're talking
#75,000 PSI neighborhood here, proof load territory.  The remaining four
#cartridges were dismantled,

This is a great data point! Thanks for risking yourself and gun :)
If our calculations are right it would be about 71K psi which shouldn't
have damaged the gun, but give a sticky bolt.   Anyone have any
opinions on that?

#
#DC Lot 63G 42.7 H335/155 Palma		5	ca. 47500 +- 2000
#
#A much more reasonable bullet weight for this powder charge
#
#
#DC Lot 64D 44.0 4895/155 Palma/RWS 5341	17	44690 SD 1051	2877 SD 30
#DC Lot 64E 44.5 4895/155 Palma/RWS 5341 18	N/A		2908 SD 19

#
#DC Lot 64A 43.0 4895/155 Palma/WLR	15	42674 SD 1038	2802 SD 24.2
#DC Lot 64B 43.0 4895/155 Palma/RWS 5341 12	42713 SD 862	2812 SD 51
#
#This test was to try and tell the difference between RWS primers and
#Winchester primers.  According to Bart, RWS primers are milder and more
#uniform than Winchester.  The expected result should be that RWS produces
#lower velocity, lower peak pressure, and more uniform velocities.
#This was not the case, in this test.  I'm interested in doing some more
#testing along these lines, I was really shocked to see a S.D. of 51 from
#the RWS primed ammo.  There was one shot that registered 2949 fps, which
#could contribute to enlarging the S.D.  Even discarding this, the SD of the
#RWS primed ammo is definitely worse.

Was the sun setting when you did the RWS primers?  That might explain
the higher SD from the chrono.
One of my few primer difference tests with .223 showed no peak pressure
difference, but some small increase in the early part of the pressure
curve with the magnum primer.  Your results show no significant difference
except in the velocity which even isn't at a real high confidence level just
guessing from the std dev of the means.

#DC Lot 64C 43.0 4064/155 Palma/WLR	12	46469 SD 1081	2864 SD 26
#
#This is an interesting one.  It is an identical charge to 64A, but it
#uses IMR 4064 powder instead of IMR 4895.  IMR4064 is longer grained,
#bulkier, cleaner burning, and one step slower than IMR 4895.  I had
#expected it to give slightly lower pressures and velocities, but the
#opposite happened.  I think this is due to the loading density being
#higher with the bulkier IMR 4064, versus IMR 4895.  The IMR4064 loads
#were not compressed, but had a very high loading density, about 98% or
#99%.  The same charge weight of IMR 4895 took up less space, merely coming
#up to the bottom of the neck.  I think the reduced ullage (free space)
#caused the pressure to peak a little higher than 4895, in spite of the
#slower burning characteristics of 4064.  All other things being equal
#(we are told), a higher loading density should be more uniform, better,
#more accurate, etc.  Perhaps 4064 should be tried for Palma loads; although
#it is difficult to meter, this is not an issue with hand-trickled loads.
#Also, it does develop full pressure, so it is burning efficiently.  And
#if it burns cleaner than 4895, fouling the barrel less, it may provide
#better accuracy on the 25th or 50th shot out of a barrel.


This is a mystery all right.  In my 30-06 the 4064 gives quite lower
pressures.  There is quite a bit of air space though.  I don't think
that the difference in the bulk density should have an effect.
If the density of the solid portion of the powders are equal, which
I believe they are, since they by all indications (from IMR specs) have
the same composition, then the air space in each load should be the same
even if one appears to fill more of the case than the other.  I'll try
measure the density of both by counting out grains and weighing them
and then computing the volume from the dimensions and see for sure.
Here is some more on the powder density effect:
At one time I thought I would enhance my internal ballistic model by
subtracting the volume of the solid powder from the case volume and then
add it back as it burned.  I then used an ideal gas equation of state for
figuring pressure.  This turned out to be worse than ignoring the volume
of the powder.  The reason is the corrections to the ideal gas formula
cancel out the effect of the added volume due to powder burning.
The first order correction to the ideal gas equation is P*(V-b)=RT
(units in terms of specific volume)
where b is the "covolume" correction factor which you can think of as
being related to the finite volume of the gas molecules.  The net equation
resembles P*(V-b+1/dens)*f=RT where the 1/dens term is the added volume due
to consumption of the powder and "f" is the fraction burned.  Early
internal ballistics work for simplicity assumed that the two terms
cancelled out.  It turns out that the covolume term is actually greater
which leads to higher pressures at higher loading densities than if you
assume the terms cancel.


#Jensen Palma Lot 13			4	48300 (3 shots)	2673 +- 28
#
#Gold Cross 150-180 is some of Jim's hunting ammo.  One of the shots in
#Jensen's indicated 56245, but had normal M.V. and no signs of pressure
#such as sticky bolt lift, etc - so it was discarded, in the generation
#of a three-shot average pressure.
#
#Very interesting result here:  Jim's Gold Cross 150 hunting (flat base
#150 grain Sierra) generated higher pressure than Jensen in my Model 70
#target rifle, but it generated much lower pressure than Jensen's in Jim's
#7600.  This is a significant result.  A few weeks later, a Canadian
#government testing lab mailed Jim a report stating that Jensen '92
#developed 57,000 psi, IVI 793 (a problem hot lot) 48,000, GC match
#43.0 45,000, GC match 43.5 48,000, GC match 44.0 49,000.
#
#This means that having target ammo tested in a SAAMI standard gun
#is probably not as good an indication of safeness as many would like
#to assume.  Target rifles are quite different from SAAMI test guns;
#just because a certain lot of ammo has been certified "acceptable" by
#an official testing agency does not ensure its safety in your target
#rifle.
#

There was an impressive difference between the pressures in your
tight target rifle and the hunting gun.  It would be interesting
to somehow measure the engraving force and bore resistance for
each gun.

#
#****RIFLE CHANGE
#
#Once again time to take out the soldering iron, and swap the harness.
#This time we attached it to rifle "269," a PAS-actioned target rifle
#equipped with a .307" Krieger barrel (about 1000 shots through it).
#
#Unfortunately, it was not possible to attach the strain gage over
#the large-diameter (0.462") portion of the chamber, as was done for
#my Model 70 and Jim's Rem 7600.  This is because the PAS action has
#a very generously sized receiver ring, that pretty much covers up
#most of the barrel that is over the main body of the cartridge case.
#
#Therefore, the gage was mounted just forward of the case mount.  This
#gave an O.D. of 1.251" (barrel diam.), and an effective estimated
#I.D. of 0.302" (basically, a guesstimated weighted average of bore
#and groove diameter).  This is a very "thick-walled" pressure vessel,
#and I was not sure how well the strain-pressure conversion factor
#would work out.
#
#In a nutshell, it didn't.  I still have some investigation to do,
#to see if I can come up with a better formula for computing this
#conversion factor.  We fired DA 65, Jensen Lot 13, IVI Mexican
#Match, GC 42.5-44.5, handloads with Olin 844 powder, 45.5 Reloder 12,
#SB 79 Mexican Match, and SB79 factory ammo.
#
#All the pressures calculated seemed too high.  Jensen Lot 13, which
#had indicated the very believable figures of 52,500 in my rifle,
#and 48,300 in Jim's 7600, indicated 64,000 units in rifle "269."
#Even taking into account the fact that the cartridge case would
#provide about 5-6000 units of support to my rifle and Jim's 7600
#(meaning that their true pressures would be in the neighborhood
#of 57,500 and 53,300, respectively), the numbers just didn't make
#sense.  Jim also tried to correlate relative pressure rankings
#between different loads in his "269" rifle, compared to the same
#loads in my rifle (our rifles have very similar chambers and
#barrels).  He tells me this was fruitless.
#

It could be that the bullet moving under the gage causes higher
strain than the gas pressure alone.  I don't know how to prove that, but
we both heard stories about how a bulge followes the bullet down the
barrel.  It could be that the squeeze due to the inertia of the bullet
mass and the base pressure causes a quite high normal force, but
I don't know how to calculate it yet.  I imagine it is similar to the
Lagrange problem in calculating the pressure gradient in the powder gas
between breech and bullet base except now there is the modulus of the
lead and the jacket to deal with.


#So, here endeth this (brief ;-) interim report.  Hope it's been interesting.
#Hmmm, I wonder if anyone has actually read this far?   ;-)

At least one:)  We'll await your filling in of some of the N/A's with
interest.  It would be nice to get data on the same loads in all of your
guns that you also had tested with the piezo gage.


Steve

From: bartb@hpfcla.fc.hp.com (Bart Bobbit)
Subject: Re: [LONG] [TECHNICAL] .308 Strain Gage Pressure Results
Organization: Hewlett-Packard Fort Collins Site

Lloyd D Reid (reidld@ecf.toronto.edu) wrote:

: We start this day testing my Model 70:

: ========Ammo======================	#shots	Pressure	M.V

: DC Lot 64D 44.0 4895/155 Palma/RWS 5341 17	44690 SD 1051	2877 SD 30
: DC Lot 64E 44.5 4895/155 Palma/RWS 5341 18	N/A		2908 SD 19
: DC Lot 64B 43.0 4895/155 Palma/RWS 5341 12	42713 SD 862	2812 SD 51

: According to Bart, RWS primers are milder and more
: uniform than Winchester.  The expected result should be that RWS produces
: lower velocity, lower peak pressure, and more uniform velocities.

I suspect the case neck tension was too tight on these loads using RWS
primers.  I've found the same type of data when bullet seating force is
greater than 15 pounds.  Note that the '92 Palma ammo used a bullet
seating force of 60 pounds.  That's also about what conventionally
full-length resized cases enable.

Another thing about getting low velocity spread with RWS primers; the
powder charge needs to be maximum.  Note the 44.5-gr. powder charge has
the lowest SD.  Next lowest charge has second highese SD; lowest charge
has highest SD.  I've never got low SD with less than max powder charges
using RWS primers.


: Jensen Lot 13 Palma			5	N/A		2892
: JB 43.0 IMR4895/155 Sierra		5	47272 SD 1809	2863 SD 15
: JB 43.5 IMR4895/155 Sierra		5	N/A		ca. 2900
: JB 44.0 IMR4895/155 Sierra		5	N/A		ca. 2950
: JB 44.5 IMR4895/155 Sierra		5	N/A		ca. 2975

: These are some of Jim's handloads, intended to clone Jensen ammo, except
: for the powder charge.  Since Jensen supposedly uses 44.8 grains of 4895,
: and all of Jim's components are identical to Jensen's, it is somewhat
: surprising to see Jim's (lighter-loaded) ammo giving higher velocities
: (and pressures, too) than Jensen.

Considering the fact that his loads did not use the same powder lot,
primer lot and loading dies/processes, these differences are a sure thing.
Velocity and pressure differences happen when different things are used.
The powder lot Jensen's ammo was loaded with was a non-canistered lot in
100-pound barrels.

BB

From: bartb@hpfcla.fc.hp.com (Bart Bobbit)
Subject: Re: [LONG] [TECHNICAL] .308 Strain Gage Pressure Results
Organization: Hewlett-Packard Fort Collins Site

Lloyd D Reid (reidld@ecf.toronto.edu) wrote:

: Thanks for the comments, Bart.  My necks are turned to 0.0122", F/L
: sized without expander ball, giving about .003" of neck grip.  I'll
: measure the seating force tonight.

Betcha two dollars it's over 40 pounds with that much of `neck grip.'
If you prefer to use an expander ball, try one that's about .3082-in.
in diameter.  That'll open the mouth up to where it's about .0007-in.
smaller than the loaded round.

: I know that different lots make for different results, but I still found
: these differences to be quite a bit larger than I thought they'd be.

Such are things with non-canistered powder lots.

: Did Jensen use non-cannister powder for:
: 	* economy?  i.e., cheapest way to buy 1500 lbs of powder
: 	* selection?  i.e., choose a faster or slower grade
: 	* uniformity? i.e., only way to know that all 1500lb are the same lot

Buzzzzzzzzzzzzzzzzzzzzzzzzzzz!  That's a buzzer sounding for wrong answer
on all three counts.

That powder was surplus from a military arsenal.  For some reason, the
arsenal didn't want it.  He found out about it, bit on it, then got it.
At least that's what I remember him telling me.

: I've also heard that Jensen ammo is faster (muzzle velocity) than it needs
: to be, now that the shooting community has a few years of experience with
: the 155 bullet.  Any comments on this statement?

I doubt if it's faster than it needs to be.  Most of the USA team members
got between 2920 and 2990 fps with it, depending on exact barrel hole
and length dimensions.  We've determined that the 155-gr. Palma bullet
needs to exit the barrel with at least 2900 fps to remain supersonic through
the 1000-yard target at sea level with the temperature at about 60 degrees F.

: Do you know how fast the
: Kiwis are loading the '95 ammo?

No, but I'll probably have an educated guess very soon.  I'm getting an
8-lb. caddy of Mulwex N2208 in a day or two and am anxious to see what
it will do.  A couple of folks have already used the 45.3-gr. charge of
it in '92 Palma cases with Win. WLR primers; muzzle velocity was about
3005 fps as I remember.  That's good, as the range at NZ is about 200
feet altitude and in January, it oft times is in the 50 degree F range.

BB

From: sfaber@intgp1.att.com
Subject: Re: More Pressure Testing Stuff

#Note that pressures generated in a SAAMI test gun are not necessarily
#what you're going to get in an individual rifle.  In the case of the
#target rifles typically used in Canadian highpower competition, the
#major difference is bore diameter.  We have found that rifles using
#a tighter-than-nominal barrel will shoot issue ball ammo much more
#accurately than a nominal .308" (bore) by .300" (land diam) barrel.
#Our barrels are typically .307" bore by .298" land, and we use a
#twist of 1 in 14" instead of the standard 1 in 12" for .308 Winc.
#
#"IVI 793" is a particular lot of Canadian service 7.62 NATO that
#gave us (Ontario Rifle Association) some trouble last summer.  It
#show definite signs of high pressure in most competitor's rifles.
#My rifle, for instance, required some effort to lift the bolt open.
#Also, the hard millitary primers showed flattening (quite unusual
#with military primers).  When shooting Jensen Palma ammo (a very hot
#load), my bolt doesn't give any hint of pressure.

Your other data on the pressure difference between your hunting
rifle and your model 70 together with this is an impressive result.
Do you have any more IVI793 to get strain measurements on with your
rifles?   It would be nice to try the same ammo from the SAAMI test
in your model 70 and the hunting gun.

#Whatever the explanation, there is something going on here.  Either the
#"sticky bolt" syndrome is bogus, or IVI 793 is generating much higher
#pressures in our competition rifles.  If the latter case turns out to be
#true, it would tend to indicate that pressure testing according to
#SAAMI methods is not particularly relevant to the safety and performance
#of target rifle ammo.  In which case, how should match ammo manufacturers
#assure their customers of the safety of their product?

Right!  Need to measure the engraving force and add this into a
good model to predict the pressures for different guns.

I doubt if there is that much difference in the hardness of the
brass.  All I tested was about the same.  Avery's "Sniper Loads" book
relates his testing of various ammo for brass hardness and he did find
some soft brassed ammo which he claimed would be bad to use for hot
loads.

#We're going to pull some bullets from IVI 793, and see if perhaps they are
#a bit oversize.  Maybe we should do a Brinell hardness test on these bullet's
#gilding metal jackets, and on the cartridge cases.  Anybody out there
#got a Brinell tester? ;-)

Yeah, send them over:)

#We also got a few pressure trace curves from the government testing.
#In most of the traces, you can see a little blip from the primer firing
#(about 2-5000 psi), followed by a short delay (perhaps 50-100 microseconds),
#then the standard exponential buildup to peak pressure, and decay.
#
#I was rather suprised to see such a low pressure generated by the primer;
#I thought it would be higher.  It makes me wonder how well their pressure
#testing setup works for low pressures (i.e., perhaps the cartridge case
#provides a lot of shielding, at low pressures).

That is interesting.  Maybe we could measure the primer pressure using
a strain gage on some kind of pressure vessel or maybe cartridge
itself.

#I was also disappointed that a rarefaction wave was not visible.  When the
#bullet exits the muzzle, the gun gas is at a pressure of 3000-10,000 psi,
#depending on case capacity, barrel length, and powder charge & type.
#As soon as the bullet clears the muzzle, the no-longer-constrained gun
#gas starts rushing out.  As a result, a rarefaction (low pressure) wave
#is sent down the barrel, toward the chamber. (In supersonic flows, and
#unsteady flow, changes in pressure or velocity are propagated by means
#of rarefaction waves, compression waves, or shock waves).  This rarefaction
#wave should cause a definite drop in chamber pressure, which should be
#measurable.  I couldn't find this on the (faxed, photocopied ;-) pressure
#trace that I was given.  Perhaps when I get the actual data points on
#floppy, something may be visible.

I've seen these as spikes on the curve sometimes, sometimes they aren't
visible, and sometimes they appear as reverse spikes.  I dont' know
why yet.

#- Daniel

Steve

From: sfaber@intgp1.att.com (Steven R Faber +1 708 979 3147)
Subject: Re: "HIGH-PWR" Loads, Powder Lots, & Testing

#(Daniel Chisolm)
#Oooboy, I'm a-goin' to be making a bit of a retraction in this
#article...  But I hear that it's good for the soul, so here goes....
#
#In an earlier article I (brdyf01@nt.com) wrote:
#>In article <2rfrcd$5qo@tadpole.fc.hp.com> bartb@hpfcla.fc.hp.com (Bart Bobbitt) writes:
#>#
#>#Another interesting thing is that when ball powder is used, the front
#>#1/3rd of the barrel tends to be hotter after N rounds are fired when
#>#compared to the same number of extruded powder rounds.  Which may
#>#explain a relationship between delayed ignition and hotter front ends
#>#of the barrel when ball powder is used.

I have a scope trace of the strain vs time curve for IMR4895 in .223 Rem
and some with Win 748.  The 748 sustained the pressure longer after the
peak, where the IMR dropped off more quickly.   A curve of AA2230 ball
powder looked more like the IMR powder though.  The AA2230 is a faster
powder than the 748, and I think it is difficult to compare the loads
without first determining which characteristics are going to be kept
constant.  If I take equal charges of a fast and slow powder, the faster
one will peak earlier and higher (pressure)  and reach all burnt earlier,
but if I increase the slow powder load it can be made to peak earlier
and reach all burnt earlier also since the pressure level has increased.

It would be nice to be able to say that ball powder reaches all burnt later
in the barrel than tubular powder and takes longer to burn and that
explains the above observations, but it isn't that easy.  There is also
the fact that most ball powder is about 10% NG and is therefore hotter
than IMR powder which has no NG.

I ran some internal ballistic simulations to see what happens when
you go from a tubular powder to a more digressive shape like cordite or
ball (which is even more digressive).   My load was 45 gr of IMR4064
with a 220 grain bullet in a 30-06.  My model used the size of the
powder grain, the force constant of the powder as obtained from IMR,
the density of the powder, the ideal gas correction (covolume ) and gamma
(1.28) both values typical for single base powders.
The Thornhill approximation was used to figure
the heat loss to the barrel, which turned out to be 39% of the bullet
energy (at any given point in the barrel) was assumed to be heat lost
to the barrel.   ( Similar models of pistol loads show only about 9% loss,
so the effect is a lot more important in rifles).  The linear burn
rate parameter was adjusted to give results that matched measured
velocities and pressures (obtained with peak strain measurements).
This model predicts the velocities of a series of different loads
of varying bullet weights 150, 168, 220 grain and different charges
to within about 50 ft/sec and predicts the pressures to within
a few K psi.  ( I have to write this up sometime.)
Anyway to get on with it, if the web thickness is held constant and
the powder is made into more of a ball shape, the result is that
the pressure peaks earlier and higher from 52K to 92K and the all burnt
point moves out toward the muzzle a bit.  The barrel time went from 1.9ms
to 1.7 ms.  If the web thickness is increased to slow the powder down, keeping
it a ball shape, so the pressure is back to 52K, the all burnt point goes out
to right about the end of the muzzle, and the peak pressure point is still
a bit earlier than the original IMR 1.3 in vs 1.7 in.  The barrel time is
still 1.8 to 1.9ms though, not much change.
All the above does not take into account the effect of retardants.
The action of the retardant would be to take the ball powder back to
more progessive burning.  Unfortunately I haven't been able to get
any data on retardants from the powder companies.  Must be trade secrets.


#Prototype lots of Gold Cross match ammunition were fired, with
#43.0, 43.5 and 44.0 grains of powder.  The powder is a non-cannister
#lot of older military ball powder, "slightly slower than IMR 4895"
#
#The two extruded powder loads showed very little delay between primer
#firing, and pressure buildup.  In fact, the primer spike was merged
#in with the main pressure buildup curve.  Jensen Lot 13 took
#0.12ms to reach 10,000 units of of pressure, and 0.42ms to reach the
#peak pressure of 65,000 units (which, when calibrated, corresponds to
#56,860 psi).  IVI C21 Ball ammo was very similar.
#
#The Gold Cross (ball powder) loads showed a definite delay between
#primer firing, and main curve buildup.  In fact, there was a very
#noticeable gap between the primer spike and the main curve.  From
#primer firing to 10,000 units took 0.29ms, and from primer firing
#to peak pressure of 57800 units (47706 psi) 0.63ms elapsed.
#
#
#The Jensen (stick powder) reached higher peak pressure than the Gold
#Cross (ball powder).  It's pressure buildup curve rose at a higher rate
#(10,000->peak in 0.30ms for Jensen, 0.34ms for Gold Cross).


This is an interesting result.  It suggests that the retardants significantly
delay the ignition time and initial burning, which would be consistent
with the theory that the ball powder generally gives longer barrel times.



#>Any others?
#
#Measure the pressure-time curve.  Double integrate, making certain
#assumptions about engraving forces and frictional losses.  Included
#one strain gage at muzzle, to get an accurate time of bullet exit -
#use this to apply a linear correction to your integration procedure.
#
#A friend of mine has offered the use of a Fluke digital Scopemeter.  With
#this, I will be able to capture such a pressure-time series (presently,
#I am only able to read peak pressure).  If there is interest, I can post
#this data.... let me know (by email) if you'd like to see this.

Yes! We could compare the P-time curve predicted by the numerical model
and adjust the model to try and duplicate the measured curve.  The model
would give the pressure-distance curve easily.  I think it would be
hard to back integrate to get the P-distance curve due to all the assumptions
about the system that enter in.

So far my model generated P-t curves seem to drop off faster than the
measured curves.  It could be that the bandwith of my scope system is
broadening the curves though, so this could use more work.


Steve Faber

Subject: Measure Chamber Pressure
From: sfaber@netcom.com (steve faber)
Date: Feb 23 1996
Keywords: pressure ballistics electronics
Newsgroups: rec.guns

You can measure chamber pressures with an inexpensive peak
strain meter.  This meter is a small self contained platic cased
unit with integral LCD display on a sloped panel.  It is powered
by a single 9V battery in an accessible battery compartment.
The meter is self adjusting and self zeroing to compensate for
temperature drift, and is easy to use.  A reset button returns the 
reading to zero and the peak strain is displayed after each shot.
A calibrate button simulates a shot and can be used to set the gain
for initial calibration purposes.  A balance/track mode is selectable
to allow static strain measurements to be taken and to set the initial
balance of the device.  There is a low battery indicator on the display.
The relative chamber pressure can be read directly, or you can convert
to an estimate of the absolute pressure with a formula using the gun's
chamber dimensions.

I have some prototypes available for $129, and I'll include
a copy of the Advanced Numerical Internal Ballistic Model
with on-line tutorial.  This program will do the pressure conversion
for you and is also handy to store your test results and compare
them to theory at the same time.

No computer is needed to use the meter, however.

Email if interested, or write the address below.

Thanks,
Steve Faber

-----------
Fabrique Scientific		chamber pressure electronics
P.O. Box 536			internal ballistic models

Subject: Re: Measure Chamber Pressure
From: sfaber@netcom.com (steve faber)
Date: Feb 27 1996
Newsgroups: rec.guns

In article <4gl5nm$1oeb@msunews.cl.msu.edu>,
Stephen Swartz <swartzst@pilot.msu.edu> wrote:
#In>
#This sounds an awful lot like Oehler's new strain-guage based chamber]
#pressure instrument (ref American Rifleman articles about 6 and 2 issues
#ago).

It is true that this strain meter uses the same technique as the Oehler
PBL system (trademarks of Oehler Research).  The method has been used
for many years, however.  There is an article in the September-October
1989 issue of Handloader that gives a schematic of a similar meter.
I built a variant of that circuit and have been enhancing it ever since.
There have been components recently released that make this type of
analog meter much more economical to produce than in 1989.
The Oehler system measures the entire pressure curve, digitizes the
signal and stores in in a laptop computer which is also required for the
system.  It includes a chronograph as well. 
One can obtain a board that fits in a PC that can with minimal external
circuitry, digitize and record such a pressure curve with even greater
resolution (12 bit vs 8 bit for the Oehler) for about $350, but then
if you add in the price of a good chrono like Oehler's you end up
with a system that costs about as much as the PBL ($600 last time I checked).
Usually the peak pressure is the data one is most concerned with, and
the Fabrique Strain Meter will record this just as well.
One advantage is that no laptop is required which makes the system
more environmentally rugged.  If you already have a chronograph, this
makes a relatively inexpensive addition.  Even if you have an Oehler
PBL, you could use the meter for easier field measurements :).

The drawback to the strain gage method, is you have to glue a small
1/4" by 1/2" film (this varies since you can get different sizes) to
the gun chamber.  The surface must be sanded down to bare metal, so
this can deface the gun.  You can put the gage under the chamber by the
stock to hide it and then run the leads out.  It is particularly easy
to attach the gages to military semi-autos.
You can also use a small connector so the lead wire can be detached
(included).

The results are worth the inconvenience.  You can plot pressure vs
charge, or velocity vs charge, pressure vs bullet mass or type.
Find the temperature dependence of your loads.  Develop loads for wildcats,
or make "Mexican Match" ammo safely.

I ran extensive tests on AR15 loads with various powders.  I'll post
some results soon, but in summary I found 27 grains of H335 to best
match the factory ammo for 55 gn ball (Win 55gn BT bulk bullets).
That powder gave practically zero temperature dependence, and could be
loaded to 28 grains withouta great increase in pressure.  Generally
most powders gave a 1.5% increase in pressure with every 10 deg F
increase in temperature.  
The exceptions were AA2460 and VV N135  which were more like 3.5%.
The effect with AA2460 was worse with the higher charges and the N135 had a
worse effect with the lighter loads.  IMR4895 gave a slight negative
pressure effect with temperature.
The load that gave the greatest velocity while matching factory pressures
was 29 gn of  WW748.  Results may vary with the powder lot or gun.
It is interesting how far off reloading manuals can be sometimes.
One manual listed similar loads for BLC2 powder and H335, but
the results showed even a full case of 29 or 30 gns of BLC2 could not
even approach factory pressure levels or velocities.