1911Forum banner

261 - 280 of 382 Posts

·
Premium Member
Joined
·
3,040 Posts
In Post #252, I stated that the hammer velocity at impact with the firing pin was 44 ft/sec. That is an error. Looking at my notes, 44 ft/sec was the firing pin initial velocity. The hammer face is moving at about 27 ft/sec at impact with the firing pin.

The collision between the hammer and the firing pin is, for all practical purposes, an elastic collision which simplifies calculations. Also, since the initial velocity of the firing pin is zero (0), the calculations are further simplified, resulting in the two following equations:

Vfh = Vih*(mh - mf)/(mf + mh)
and

Vff = 2*mh*Vih/(mf+mh)

Where:

Vfh = Final velocity of the hammer
Vih = Velocity of the hammer at impact with the firing pin
Vff = final velocity of the firing pin
mh = mass* of the hammer
mf = mass* of the firing pin

*Any units of mass or weight may be used (lbs, oz, kg, gn, etc) as long as the units are the same for the hammer and firing pin.

The weight of a steel government firing pin is about 70 gn (grains). The weight of a steel government hammer is about 300 gn. The final velocity of the steel firing pin is about 44 ft/sec as measured experimentally using a steel government hammer powered by a 23lb rated mainspring.

Therefore, solving the equation for the final velocity of the firing pin:

Vff = 2*mh*Vih/(mf+mh)

For Vih (velocity of the hammer at impact with the firing pin) yields:

Vih = Vff*(mf + mh)/(2*mh)

Substituting the values from above gives:

Vih = 44*(70 +300)/(2*300) = ~27 ft/sec

And:

Vfh = Vih*(mh - mf)/(mf + mh)
= 27*(300 - 70)/(300 +70)
= ~16 ft/sec

For the steel hammer and firing pin. The hammer was theoretically slowed by about 11 ft/sec by hitting the steel firing pin. However, in real life, since the hammer is not merely traveling alone at collision, but is being powered by the mainspring, the hammer is not significantly slowed.

Having the velocity of the hammer at impact with the firing pin, we can now use any sets of values for the masses of the hammer and firing pin and calculate the final result of each pair. Alternately, the components can be altered (mainspring weight, firing pin weight, and/or hammer weight), the resultant firing pin velocity measured experimentally and the values for hammer velocity (before and after impact with the firing pin) calculated.

One can easily substitute the lower weight of 39 gn for a titanium firing pin in the equations above and see that the titanium firing pin initial velocity is only about 4 ft/sec higher than the 70 gn steel firing pin.

Only by using a lighter hammer (about 70% the weight of a government steel hammer) with a 23lb rated mainspring can the titanium firing pin have enough velocity to give it the energy of the government parts to ignite the primer.
 

·
Premium Member
Joined
·
14,760 Posts
I wonder if JB thought of all of this.
Without quoting history designs are a compilations of past experiences with a few new ideas or approaches in some areas, but undoubtedly science and the factors are calculated as the design takes place. When Colt won the contract they then modified the patented design to their manufacturing interpretation. As we know the thumb safety wasn't part of the 1911 patent and was added with the 1913 patent which when examined has a variety of patented ideas that were not ever actually used as per the patent. The thumb safety in the patent also locks the hammer, but would have been an added fitting problem so only the sear is blocked.

Walt Kuleck must have read the patent and thinking the thumb safety features were implemented claimed in his book The M1911 Complete Owners Guide, that the thumb safety would indeed block the hammer, and actually claimed in his book that he always checked this on every one he worked on, only to be challenged and finding he was very much mistaken. Colt did not follow this patented design, nor any of the other 1913 patents, such as a strut that keyed into the hammer, not needing a pin, and grips that slid on a key hole and were locked in with the MSH pin. Is kind of an interesting doc to examine.

https://patentimages.storage.googleapis.com/d8/a5/5c/1722057992bc87/US1070582.pdf

LOG
 

·
Registered
Joined
·
539 Posts
What we define as the reference and how we know it with accuracy I think is dependent on the jig we use to create the sear's flat. I looked at a lot of jigs. All provided no detail on principle of operation. Pictures and youtube video helped a little, but not much. Instructions were basically 4 clicks here, 20 clicks there, what does a click even mean? I settled on a simple Ed Brown Sear Jig because it was clear how it functioned and I could see how to make it more accurate. I'll try to provide thoughts on this in a couple of days.
I modified a Brownells sear fixture to allow precise control over the sear face angle.

As supplied, the "adjusting screw" was frozen in position with a threadlocking compound. This screw located the lower feet of the sear in the fixture. The sear tip would then presumably be in the correct position. Freeing up the screw allowed adjustment, but the screw is only accesible if the sear is swung out of position. So it is trial and error to position the sear.

The modified fixture has an adjusting screw that's accessible while the sear is in position. It also has a tensioning spring to hold the sear in position against the adjusting screw. I have a machined face on the fixture with a precisely known distance from the sear axis center. Measuring from this face, I can locate any part of the sear face directly over the sear axis. A measurement of the sear tip protrusion above the fixture top provides a value for a shim to rest the stone on. The stone is thus parallel to the fixture top surface and the sear tip, and perfectly at right angle to the vertical line from the sear axis.

Friction (and the forces and moments is creates) has a directionality. It's always opposite to the imposed direction of movement. That is friction creates hysteresis in force analysis. When making a measurement (where friction plays an important role) we want to make sure the direction of movement is in the direction the system will actually have for the thing we seek to measure. We know when we pull the trigger the sear will move from under the hammer hooks and the hammer will fall. Thus when we measure the hammer's force on the sear we want the hammer's direction to be falling.
I understand the friction and hysteresis. I guess I don't quite understand how you are measuring the torque?

I had assumed that we were both starting from the same initial condition, with the hammer at full cock. And we are both pulling in the same direction. The only difference is that I am using a spring scale and you are using weights.

Anyway, I can try using your weight method.

Friction is the wild card. I would love to directly measure it. But I'm not yet sure how to do it. I'm all ears to your ideas here.
As friction would apply to a particular steel and surface, you could measure the static friction at least, using a true radius sear and a trigger pull gauge.
The hook force can be derived from hammer torque measurements, and the normal can be calculated from that using the part geometries. From that, you can calculate the static coefficient and apply it to other sear geometries of the same material and surface finish.

There are some things I am not sure about, but perhaps you may know:

Can the friction force and a cam force be simply added as force vectors? The two vectors are along the same axis, and so would the composite force be a simple sum?

If so, the static friction for non-radiused sears could also be determined by calculating the cam force and subtracting it from the measured disengagement force.

Also, do the static and sliding friction coefficients track each other? If so, you may be able to just calculate the sliding friction from the static friction. I can't think of any practical way to measure the sliding friction over the tiny lengths of travel involved without specialized instruments.

One way to do it might be with a load cell and a carefully controlled pull motion. The requirement is a pull force with very high spring rate. A spring pull gauge is too reactive, and will either snap the sear out of engagement, or allow it to move in small jumps to new points of static equilibrium.

With a load cell based gauge, the sear's motion would closely follow the gauge's motion. The pull would be continuous and unyielding and at the same time, the force would be registered.

I believe digital pull gauges use load cells, but you would still need a mechanism for a precise continuous pull, and I don't believe there would be any way to access the data registered during the pull.

-
 

·
Registered
Joined
·
539 Posts
I need to see a picture of what you are describing to fully understand your thought here. This is my crude attempt to picture it. Let's use your wedge concept.

A 90 deg sear at trigger break is like a wedge with zero angle. All one sees here is friction. The normals present but internal to the system (something like circulating torques in a gear box) and don't alter trigger force. A 92 deg sear is then like a 2 deg wedge. The wedge under no friction would just squirt out. It's friction that holds the wedge in. As the wedge becomes steeper friction can no longer hold it in, it squirts out. Is this somewhat along the lines of your thinking?
Yes, that is the basic principle.

My method of analysis for the cam force uses simple mechanics, and goes straight to the end results. A hammer hook, under a certain tension, moves by a certain amount, while a sear moves by a certain amount. That's everything we need to know. Everything involved is incorporated into this result.

This illustration may help:

611756


The calculations and the terms would be as follows:

F_hook is the hook force as derived from hammer torque measurements.

alpha is the hook displacement angle relative to the sear axis location.

dx_sear is .0001". Results are more accurate as this term becomes smaller. But the size is limited by the precision of the math functions.

x1_sear is the point along the sear face that is directly beneath the hammer hook tip. It is varied in .001" increments from the front edge of the sear to the escape edge. This point is relative to the 90 degree vertex point.

x2_sear = x1_sear - dx_sear

y1_sear = square root(x1_sear² + .4055²)

y2_sear = square root(x2_sear² + .4055²)

dy_sear = y1_sear - y2_sear

dy_hook = dy_sear / cosine(alpha)

F_cam = F_hook * cosine(alpha) * dy_hook / dx_sear

-
 

·
Registered
Joined
·
96 Posts
Discussion Starter · #267 ·
In Post #252, I stated that the hammer velocity at impact with the firing pin was 44 ft/sec. That is an error. Looking at my notes, 44 ft/sec was the firing pin initial velocity. The hammer face is moving at about 27 ft/sec at impact with the firing pin.

The collision between the hammer and the firing pin is, for all practical purposes, an elastic collision which simplifies calculations. Also, since the initial velocity of the firing pin is zero (0), the calculations are further simplified, resulting in the two following equations:

Vfh = Vih*(mh - mf)/(mf + mh)
and

Vff = 2*mh*Vih/(mf+mh)

Where:

Vfh = Final velocity of the hammer
Vih = Velocity of the hammer at impact with the firing pin
Vff = final velocity of the firing pin
mh = mass* of the hammer
mf = mass* of the firing pin

*Any units of mass or weight may be used (lbs, oz, kg, gn, etc) as long as the units are the same for the hammer and firing pin.

The weight of a steel government firing pin is about 70 gn (grains). The weight of a steel government hammer is about 300 gn. The final velocity of the steel firing pin is about 44 ft/sec as measured experimentally using a steel government hammer powered by a 23lb rated mainspring.

Therefore, solving the equation for the final velocity of the firing pin:

Vff = 2*mh*Vih/(mf+mh)

For Vih (velocity of the hammer at impact with the firing pin) yields:

Vih = Vff*(mf + mh)/(2*mh)

Substituting the values from above gives:

Vih = 44*(70 +300)/(2*300) = ~27 ft/sec

And:

Vfh = Vih*(mh - mf)/(mf + mh)
= 27*(300 - 70)/(300 +70)
= ~16 ft/sec

For the steel hammer and firing pin. The hammer was theoretically slowed by about 11 ft/sec by hitting the steel firing pin. However, in real life, since the hammer is not merely traveling alone at collision, but is being powered by the mainspring, the hammer is not significantly slowed.

Having the velocity of the hammer at impact with the firing pin, we can now use any sets of values for the masses of the hammer and firing pin and calculate the final result of each pair. Alternately, the components can be altered (mainspring weight, firing pin weight, and/or hammer weight), the resultant firing pin velocity measured experimentally and the values for hammer velocity (before and after impact with the firing pin) calculated.

One can easily substitute the lower weight of 39 gn for a titanium firing pin in the equations above and see that the titanium firing pin initial velocity is only about 4 ft/sec higher than the 70 gn steel firing pin.

Only by using a lighter hammer (about 70% the weight of a government steel hammer) with a 23lb rated mainspring can the titanium firing pin have enough velocity to give it the energy of the government parts to ignite the primer.
You put a lot time into thinking about this and I seek to respect your effort in my comments. If that doesn't come across then please let me know.

I believe your analysis is at an infinitesimal time just before the hammer strikes the firing pin and just after. You are assuming an elastic collision, therefore conservation of energy (here kinetic energy) is applied. All this is reasonable.

So we begin with, KE_hammer_initial= KE_hammer_final+KE_fire_final. Where KE= 1/2 m v^2

I'll accept your value of v_hammer_initial as correct.

The kinetic energy equation has two unknowns, v_hammer_final and v_fire_final.

So we turn to conservation of momentum (momentum= velocity *mass). Here m_hammer*v_hammer_initial= m_hammer_final*v_hammer_final + m_fire_final*v_fire_final.

We now have two equations and two unknowns and can solve for v_hammer_final and v_fire_final in terms of the value for v_hammer_initial.

As we are working across two different quantities, energy and momentum, we need to have the correct units for mass and velocity. The easiest to use here is metric, so mass is in kg and velocity in m/s. Once solved for in these units one can convert to units of your choice.

Here is where I depart from your analysis.

I don't see any square root terms in your equations. As KE is velocity squared I would expect to see square roots when expressing final velocities for the hammer and firing pin in terms of initial hammer velocity.

Also the model for hammer and firing pin used here are two masses which traveled in a straight line before and travel in a straight line after collision. This means that the center of mass of these two bodies are moving in a straight line and thus those centers' are displaced, hence the full effect of their respective masses are involved. This is true for the firing pin, but not the hammer. The hammer rotates (about its pin) this must be accounted for. Additionally, the center of mass of the hammer is not located at its contact face with the firing pin. This means that a reduced value for the hammer's mass must be used if one uses a straight line motion analysis.

You can get a solution, its just a little more complicated.
 

·
Premium Member
Joined
·
8,664 Posts
Not to be knit pickie but how complicated does it need to be? One could also apply the peripheral speed of the effective mass at the contact point, its exact vector application, etc., but the issue is, will it cause primer ignition?

While there has been some good discussion 75% of this thread has been not seing the forest for the trees.

Carry on
 

·
Registered
Joined
·
96 Posts
Discussion Starter · #270 ·
I modified a Brownells sear fixture to allow precise control over the sear face angle.

As supplied, the "adjusting screw" was frozen in position with a threadlocking compound. This screw located the lower feet of the sear in the fixture. The sear tip would then presumably be in the correct position. Freeing up the screw allowed adjustment, but the screw is only accesible if the sear is swung out of position. So it is trial and error to position the sear.

The modified fixture has an adjusting screw that's accessible while the sear is in position. It also has a tensioning spring to hold the sear in position against the adjusting screw. I have a machined face on the fixture with a precisely known distance from the sear axis center. Measuring from this face, I can locate any part of the sear face directly over the sear axis. A measurement of the sear tip protrusion above the fixture top provides a value for a shim to rest the stone on. The stone is thus parallel to the fixture top surface and the sear tip, and perfectly at right angle to the vertical line from the sear axis.
Excellent... You saw the same problems with the Jig as I and fixed them. The Ed Brown Jig is similar with similar problems. You saw the supplied shim as wrong (it needs to be 0.035" for Ed Brown not 0.020") and something different is needed to ensure the stone is parallel to the long top flat of the jig. The tensioning spring is a clever adder. Screws are terrible positioning devices unless loaded as there is slop in threads (both vertical and horizontal). The precisely machined face you are using for a reference to exactly locate the escape edge above sear pin center is spot on. I'm doing the exact same thing.

With those fixes I think the jig will be very accurate and highly repeatable.

So here is a funny. The Ed Brown Jig states the set screw should a specific distance, +/- 0.001. One finds in solids modeling, with an ordnance sear, that specific distance places the escape edge directly above the sear pin sear. Great. But nobody uses ordnance sears with a sharp point at the escape edge. Nearly everybody cut the 45 deg relief. So the sear angle that is then created by the jig is greater then 90 deg. Maybe that is the intent, or maybe they don't understand.

I understand the friction and hysteresis. I guess I don't quite understand how you are measuring the torque?

I had assumed that we were both starting from the same initial condition, with the hammer at full cock. And we are both pulling in the same direction. The only difference is that I am using a spring scale and you are using weights.

Anyway, I can try using your weight method.
I don't measure the torque, I calculate it from the measured weight. So there is some uncertainty in the torque calculation due to uncertainty in the moment arm. Using a square I line up one edge of the square with the vertical wire holding the weight. I raise the square up till the other leg of the square is cutting the center of the hammer pin. I now know the length of the weight's moment arm. To gain repeatability, the pistol is held in a vice. The vertical wire is just a little beyond the bottom edge of the pistol's backstrap at full cock. This way the moment arm will be equal from experiment to experiment.

As friction would apply to a particular steel and surface, you could measure the static friction at least, using a true radius sear and a trigger pull gauge.
The hook force can be derived from hammer torque measurements, and the normal can be calculated from that using the part geometries. From that, you can calculate the static coefficient and apply it to other sear geometries of the same material and surface finish.

There are some things I am not sure about, but perhaps you may know:
That is a clever idea. It may be a way to start. I just get nervous swapping in a different sear as the effect we measure could be due to may other things other then friction and we won't know. But again it may be a start.

Here is a funny. My sear and hammer hooks have been untouched since I adjusted them well over a month ago. I didn't want to go missing with them yet as results I have collected then needs to be fully repeated. With just CLP to lube the leafless trigger break was about 3/4 lb. Fully cleaning and applying moly dry power the break dropped to about 1/2 lb. These measurements were always with the same "23#" mainspring. I just swapped in my "19#" mainspring. I now need to use weights to accurately measure break as it is under 5 oz. I think it best if we use a "23#" mainspring when we talk, otherwise we'll have apples and oranges.
Can the friction force and a cam force be simply added as force vectors? The two vectors are along the same axis, and so would the composite force be a simple sum?

If so, the static friction for non-radiused sears could also be determined by calculating the cam force and subtracting it from the measured disengagement force.

Also, do the static and sliding friction coefficients track each other? If so, you may be able to just calculate the sliding friction from the static friction. I can't think of any practical way to measure the sliding friction over the tiny lengths of travel involved without specialized instruments.
Vectors can always be added together. I need to see diagrams of your analysis before I can comment further. You may have shown them in the next post you made.

When you ask if the static and sliding coefficients track I assume you mean do they have a fixed proportionality between them. The answer is generally yes, but we may not be able to discover that constant. Also I hate to think the sliding coefficient might be velocity dependent. That is a different value for the bullseye shooter from the action shooter. Specialized instruments is likely beyond our means. In the end we may have to settle for a bounded solution rather then a fairly specific number.

One way to do it might be with a load cell and a carefully controlled pull motion. The requirement is a pull force with very high spring rate. A spring pull gauge is too reactive, and will either snap the sear out of engagement, or allow it to move in small jumps to new points of static equilibrium.

With a load cell based gauge, the sear's motion would closely follow the gauge's motion. The pull would be continuous and unyielding and at the same time, the force would be registered.

I believe digital pull gauges use load cells, but you would still need a mechanism for a precise continuous pull, and I don't believe there would be any way to access the data registered during the pull.
Within this thread someone mentioned a trigger measurement system which had a loadcell, likely a ball screw, and stepper motor to move the trigger in a controlled matter. The system continuously recorded force and displacement. I have designed and built such devices in the past. This one is German made, which generally means it is high quality. However, for a German product there is a surprising lack of any performance information. The one spec they provide states they measure (sense?) force to 1 part in a million (re full scale). Thus their claimed signal to noise ratio is 120 dB. This is basically impossible to achieve in a mechanical system.

You are on target thinking the applied load needs to have a high stiffness, ie a high spring constant. Basically one needs a sensor with a stiffness at least 10x higher then stiffness of the system one is seeking to gain information on. Otherwise the deflection (compression) of the sensor becomes entangled with the system's deflection that one is seeking to measure.

Loadcells can be stiff, but cheap ones aren't as they cost more. As you point out we need a way to measure both force AND displacement simultaneously. The displacement part is tricky. To ensure accurate displacement measurement one doesn't want any of the systems load force to pass through the displacement sensor. That is load force must be shunted around the displacement sensor. I don't think the German design does this. They likely said we have a ball screw, it has so many turns/inch. We have a stepper motor, we can easily count teeth. We now have a displacement sensor. Yeah.... but it won't accurately resolve 0.0001" which is what we need.
 

·
Registered
Joined
·
5,139 Posts
I love seeing engineers picking apart a gun designed by man who didn’t even consider himself a proper engineer.

Let me ask the group this. Does any of this discussion make trigger work easier? Does any of this advance the craft? Is any of this information really new?

What is the main goal of this discussion and is there one?
 

·
Premium Member
Joined
·
14,760 Posts
]
I love seeing engineers picking apart a gun designed by man who didn’t even consider himself a proper engineer.

Let me ask the group this. Does any of this discussion make trigger work easier? Does any of this advance the craft? Is any of this information really new?

What is the main goal of this discussion and is there one?
Lets see, no, no, no, and self-aggrandizing, yep, that's about it.:p
 
  • Like
Reactions: Jolly Rogers

·
Premium Member
Joined
·
8,664 Posts
EA, I’ve garnered where the different points on the sear face are relative to the sear pin centerline. mega discussed that and it was an ah-ha moment for me. Otherwise, much of the information was discussed before but in an unengineered format.
 

·
Registered
Joined
·
56 Posts
Let me ask the group this. Does any of this discussion make trigger work easier? Does any of this advance the craft?
As an aerospace machinist and TIG welder, I agree with EvolutionArmory on making things easier and repeatable. I appreciate Engineers and all of their knowledge, but it comes down to butchering metal to make tooling and .0001" type tolerancing costs lots of $$$. This thread is a fun read and I hope this is accepted information from a shop floor guy and bullseye competitor.
My fix on the Marvel Jig was to make a new sear block and use a depth mic to set the 90 degree sear face to the centerline of the sear pin as shown, easy. It can be changed in .001" increments until the "feel" this "trigger princess" likes.
Chuck Warner's radius sear is much, much easier - no relief cut needed, just stone the hammer hooks down to desired feel (Bullseye only).
More pics of the 90 degree sear jig modifications.
Set the sear to the sear pin centerline and proceed to roll the fixture back and forth like a surface grinder. These pics are with an old beater sear.
611811

611812
 

·
Premium Member
Joined
·
14,760 Posts
That is a great process WES, I found the Power Custom to be very easy to register the angle desired by simply using a quality square and a stone, raise or lower the stone rest to adjust. I think I already posted this in this thread, but too many pages to search. lol I lay the stone on the sear pin before I start to assure the roller and pin is in the same plane. I also lower the blade against against the sear pin and set so the sear tip is .055" away. To hold the sear while stoning I do as Dave Berryhill did and shared. Use a hammer!


611819


611820
 
  • Like
Reactions: Jolly Rogers

·
Registered
Joined
·
96 Posts
Discussion Starter · #278 ·
Not to be knit pickie but how complicated does it need to be? One could also apply the peripheral speed of the effective mass at the contact point, its exact vector application, etc., but the issue is, will it cause primer ignition?

While there has been some good discussion 75% of this thread has been not seing the forest for the trees.

Carry on
This is an open forum. Everyone can voice their opinion. But if I was only finding 25% of what I was reading as having some value I would stop reading this thread and do something of greater value with my time.
 

·
Premium Member
Joined
·
8,664 Posts
Reading it is how I assess it. I said I picked up perspective from it. I never said to stop. Keep posting, no problem.
 
  • Like
Reactions: log man

·
Registered
Joined
·
96 Posts
Discussion Starter · #280 ·
As an aerospace machinist and TIG welder, I agree with EvolutionArmory on making things easier and repeatable. I appreciate Engineers and all of their knowledge, but it comes down to butchering metal to make tooling and .0001" type tolerancing costs lots of $$$. This thread is a fun read and I hope this is accepted information from a shop floor guy and bullseye competitor.
My fix on the Marvel Jig was to make a new sear block and use a depth mic to set the 90 degree sear face to the centerline of the sear pin as shown, easy. It can be changed in .001" increments until the "feel" this "trigger princess" likes.
Chuck Warner's radius sear is much, much easier - no relief cut needed, just stone the hammer hooks down to desired feel (Bullseye only).
More pics of the 90 degree sear jig modifications.
Set the sear to the sear pin centerline and proceed to roll the fixture back and forth like a surface grinder. These pics are with an old beater sear.
View attachment 611811
View attachment 611812
Thanks for the reply. Your radius sear has no relief cut and as a bullseye shooter you like the trigger. That's a good piece of info.

My thought was a relief cut was needed. Not just to set sear engagement length re. hook height but maybe to cam out the sear as the sear's escape edge passes under the hammer hook tips. I also thought today's hammers might need that relief to ensure the sear wasn't recaptured by the half cock notch as the hammer dropped. This issue could be more pronounced for a bullseye shooter over an action shooter as the bullseye shooter has a much slower trigger pull.

Thanks again for the info it was very useful.
 
261 - 280 of 382 Posts
Top