Reverse Engineering a ballistic program.

Discussion in 'Rifles, Bullets, Barrels & Ballistics' started by Jeff In TX, Jul 11, 2007.

  1. Jeff In TX

    Jeff In TX Well-Known Member

    Messages:
    408
    Joined:
    Aug 10, 2003
    In my real life job, I’m an engineer. I don’t like to accept almost or close enough when it comes to results. I’m anal about results and I strive to achieve exact results and once I dig into something it’s hard to let go. Sometimes my quest to achieve exact results takes unconventional methods.

    I’ve owned many ballistic programs over the years, most of them ya’ll swear by. Most in my opinion are border line adequate at best for long range calculation and estimation. I learned a lot of the ballistic formulas and even tried my hand at reworking Dave Kings excel program he sent me many years ago. But in the end I couldn’t get them to match my actual shooting data. Four years ago I set out on a quest to map my .308 actual data to my ballistic program or at least get as close as possible.

    You can agree or disagree with me on this, but in the end I accomplished what I set out to do and in the process learned a bunch.

    First if you’re using the published G1 BC from the bullet manufacture and program running the G1 drag model, you’re gonna waste a lot of time and effort trying to do this.

    The G1 drag model is based on a 100+ year-old calculation using a 1 inch 1 pound round nose projectile as its basis. So what the bullet manufactures are really saying is…. any type of bullet regardless of shape can be calculated using the same mathematical model. This is total BS. Last time I looked a VLD bullet is more aerodynamic than a flat based bullet.

    Those are the real issues I see when it comes to BC. Boat tail bullets us a G5 drag model and VLD bullets use a G7 drag model. Both drag models have a lower BC both totally different calculations for determining the bullets exterior ballistic tables.

    Okay let’s get started. When I say I reversed engineered my ballistic program, what I’m really saying is I found the proper input data and messaged it a bit to get my exact results.

    I started this in January 03. While at the TACPRO range I established a tremendous amount of shooting data. With two CED Millennium chronographs back to back at 7’ from the muzzle I shot 10 rounds to get my actual muzzle velocity. Both CED’s were within 3 fps of each velocity measurement. That’s consistency! I recorded air temps, barometric pressure, and humidity with my Krestel 4000 weather meter. I don’t deal with or use altitude. I then set one chronograph at exactly 100 yards plus or minus an inch or so and shot 5 shots and recorded my velocity. I did this again at exactly 200 and 300 yards. And yes, it’s easy to shoot through a chronograph at 300 yards and not shoot the screens.

    Also during this 200 and 300 yard process I never adjusted my scope elevation. I did however record my exact bullet drop at each of the yardages from my POA to POI.

    I then went to 400 and 500 yards and shot for group and recorded how much my bullet dropped from my POA to POI. At 600 and 700 yards it required mounting a 12 foot 2X4 for aiming purposes to measure my POA to POI. I now had exact bullet drops to 700 yards with all the environmental conditions to go with it. 800, 900 and 1000 yards I had to rely on using mildots for hold overs and then calculating my POA and POI.

    Using my RSI ballistic lab program I easily figured out which CD model to use for my bullet and what the exact CD model BC was for that bullet.

    My ballistic program results were wacky. I could get my 100 and 1000 yards to match, but the intermediate ranges were skewed, or I could get some of my intermediate ranges but my short and long ranges were skewed. I tweaked velocity and BC’s numbers and could get a tad closer but my results were like a balance beam. Adjust this side and the other end goes array.

    I had measured my scope height very accurately, but I was using 20 moa bases. Could this change my results… big time! I started to message my scope height, BC and muzzle velocity, though muzzle velocity only slightly, within 25 fps. My ballistic data from my program began to match my actual shooting data all yardages. It took a lot or tweaking, but I got the right combinations of numbers. I was within .25 moa off all 100 yard increments.

    My next trip to TACPRO was in July and it was 90 degrees already at 8:30 in the morning. I started to record all new data, but a funny thing was happening. I adjusted for air temps, barometric pressure and humidity on my RSI program and my actual results were matching exactly. I backed up from 400 yards to 1000 yards and adjusted my scope per my programs data. 5-shots within a 10” circle from POA. Holy crap it was working. When I used the ballistic data at the other ranges they were dead on.

    Later that fall I went to Colorado and was about 8000 feet in elevation and temps were in the teens. The longest we could shoot was about 750 yards. I ran the program with the new temps and barometric pressure, figured out the wind and was dead on at all ranges according to my program.

    I had cracked the code at least for my .308. BTW I tried it with the G1 drag curve and could never get close.

    That’s it in a nut shell, there is a bit more to this, but it all comes to down getting exact data at all ranges and messaging the numbers.

    You can agree or disagree, call me nuts, but I managed to achieve what I set out to.

    Let the flood gates open.
     
  2. CAM

    CAM Well-Known Member

    Messages:
    385
    Joined:
    Jan 2, 2002
    Jeff
    Good post, thats what I think a lot of us do. get set points and known data and massage the numbers to match.

    Like you state the drag model is set on most programs and your left with adjusting the BC at intervals.

    I'm a welder and we make things THEN have the Engineer draw them, No change orders! everything fits! works every time /ubbthreads/images/graemlins/smile.gif /ubbthreads/images/graemlins/smile.gif

    What drag model is the Exbal program? anybody know?

    CAM
     

  3. royinidaho

    royinidaho Writers Guild

    Messages:
    8,853
    Joined:
    Jan 20, 2004
    Jeff,

    Damn! I had an inkling that you were going to say what you said. Jeeze, more work for me. /ubbthreads/images/graemlins/frown.gif

    Regarding scope height. I'm all comfy as I'm carrying the numbers out 3 places /ubbthreads/images/graemlins/wink.gif but after ready you post I'm wondering if I'm using the right numbers.

    What's a better method of determining the correct number especially if using the plastic burris inserts to get a few more MOA?

    Great post.
     
  4. Jeff In TX

    Jeff In TX Well-Known Member

    Messages:
    408
    Joined:
    Aug 10, 2003
    [ QUOTE ]
    Jeff,

    Damn! I had an inkling that you were going to say what you said. Jeeze, more work for me. /ubbthreads/images/graemlins/frown.gif

    Regarding scope height. I'm all comfy as I'm carrying the numbers out 3 places /ubbthreads/images/graemlins/wink.gif but after ready you post I'm wondering if I'm using the right numbers.

    What's a better method of determining the correct number especially if using the plastic burris inserts to get a few more MOA?

    Great post.

    [/ QUOTE ]

    When it comes to scope height and tapered bases it's a crap shoot on what number to use. I found out the actual measured number was the problem. But when I started to make changes, everything fell into place. Afterwards I figured it was the angle/tapered bases screwing up my results.

    Ballistic programs work on basis of the scope base being parallel to the bore. I would have never guessed it, but it made a huge difference. Trial and error is my best advice!
     
  5. RockyMtnMT

    RockyMtnMT Official LRH Sponsor

    Messages:
    3,029
    Joined:
    Mar 25, 2007
    Jeff,

    Is there a formula to figure out bc using velocity @ muzzel and velocity @ 100yrds? I have some bullets that I can not find a bc value for.

    Thanks for your input.
    Steve
     
  6. Jeff In TX

    Jeff In TX Well-Known Member

    Messages:
    408
    Joined:
    Aug 10, 2003
    [ QUOTE ]
    Jeff,

    Is there a formula to figure out bc using velocity @ muzzel and velocity @ 100yrds? I have some bullets that I can not find a bc value for.

    Thanks for your input.
    Steve

    [/ QUOTE ]

    Steve,

    The answer to your question is yes. I use the RSI ballistic program because it's simple to input. I'd have to dig up the formula's but there not that hard. If you send me an email with the bullets and velocities, I'd be glad to run them for you.
     
  7. britz

    britz Well-Known Member

    Messages:
    1,217
    Joined:
    Mar 11, 2007
    JMB calc has a program online that will do this for you too. But that would probably go against your thoughts on using tradjectory programs too lol!
     
  8. Jeff In TX

    Jeff In TX Well-Known Member

    Messages:
    408
    Joined:
    Aug 10, 2003
    Britz,

    I worked and talked with Brad many times over the years (maker of JBM online). He actually had a great ballistic program for PC ("On Target", if I remember correctly) back in the 90's, but went to an online version instead.

    Brad's a great guy to talk to and work with.
     
  9. Gustavo

    Gustavo Writers Guild

    Messages:
    131
    Joined:
    Jun 12, 2007
    Jeff,

    Nice review on your experience and very valuable especially for those who don't take the time to see what's going on when data doesn't match to actual field results.

    However, some things I'l like to recall, since I posted them many times before in the past.

    1) You cannot fudge numbers without paying the consequences. "Bending the curve" will not made up for an inaccurate ballistics engine. How many times will you have to repeat the process for any new load ?

    2) Good input will make for good outputs...and the other way around.:confused:

    3) The basic problem is not the use of G1/G5/G7 BCs. They are what we have at hand, what the manufacturers turn out. The problem lies in the ballistics predictive model most programs uses, especially at longer ranges when bullets go through the transonic range.

    Not my idea to start a thread ( fight ?? :)) on ballistics models, but fortunately enough we have new mathematical tools that yield very accurate results at extended ranges.