With the characteristic slight loss of hearing from years on the firing range, and a near-permanent half-squint in one eye for focusing, Clyde Rose is every inch a rifleman. So, no one would have been surprised if he clustered his shots in one-inch groups one day while target shooting.

But half-inch groups? That was a surprise. Yet, that's what he shot when he strapped his Browning BOSS (Ballistic Optimizing Shooting System) on his Browning A-bolt rifle and began test firing with factory ammo. In later tests, one shooter actually achieved a 0.067-inch cluster. Rose's invention was clearly a winner. But how did it work?

It worked by altering the frequency of the barrel vibrations. Every time a rifle fires, the barrel vibrates in all directions. Bullet type and weight, rate of primer ignition, barrel configuration, and other characteristics influence that vibration.

Of course, Rose, Browning's chief design engineer, and his team knew that, but had no technical documentation to back up their assumption that the BOSS altered the frequency of the vibrations. That's because they had developed the BOSS by hand through trial and error. Yet, they needed documentation and verification so they could optimize the design without the time and expense of field trials.

To verify the results, and to show how software can improve the design process, Structural Research and Analysis Corp. and Parametric Technology Corp. volunteered to collaborate with Browning in a special Design News project to investigate the structural dynamics of the BOSS.

For no remuneration except the satisfaction of showing engineers how to improve their productivity, the software companies replicated the mechanics of the BOSS on their computers to verify its operation and provide a map for design improvements. Their work provided the documentation Rose and his team needed to begin design optimization.

Zeroing in on the problem. Work began in a brainstorming meeting at Browning headquarters with Rose; Raymond Allen, Browning's vice president for R&D Larry Nelson, Browning's chief engineer; J.D. Brookhart, account manager, and Michael Briscoe, applications engineer, both from Para- metric Technology Corp.; Roni Plachta, applications engineer from Structural Research and Analysis Corp.; and Design News. The first step was to model the existing BOSS in PTC's Pro/ENGINEER.

The BOSS consists of a weight, a body, and a micro-adjustable locknut. It screws onto the barrel, making a rigid connection. It adjusts the timing of the barrel's vibrations so the bullet leaves the muzzle at the same, optimal point in the barrel's oscillation, consistently. Gas-vent holes drilled at a precise angle in the body of the BOSS help reduce recoil 30-50%

The BOSS can move up and down a rifle barrel 10 mm. There are ten different settings. Adjusting the settings on the weight allows fine-tuning of barrel vibrations.

The software companies based their modeling and analysis on the effect of the BOSS on an A-Bolt II Stainless Steel Stalker rifle firing 300 Win. Mag caliber bullets. The barrel is stainless steel. The stock is 20-30% glass-reinforced polypropylene, with a Young's Modulus of 2.4 x 10⁹ Pascals. The recoil pad is rubber, with a Young's Modulus of 6.1 x 10⁶ Pascals.

SRAC Technical Support and Development Engineer Suchit Jain was project manager for the analyses. "I started with the assumption that if I could prove that the BOSS reduces the wide deflection of the barrel at the tip, I would be showing that the bullet leaves at less of an angle, thus improving accuracy," he says.

Aim for flexibility. PTC's Briscoe started the project by developing within Pro/ENGINEER a 3-D model of the existing rifle barrel and BOSS. "I wanted to make the design as flexible as possible," he says, "and the feature-based parametric nature of Pro/E made that easy."

To model the barrel, he used a command to create a revolved protrusion. He made every aspect of the model a parameter, including the angle of the holes in the BOSS. "That meant SRAC or Browning could click on any dimension and update it to make changes," he says.

When Briscoe had the parts modeled, he went into an assembly mode in Pro/ENGINEER to put all the components together using intelligent assembly constraints. "Because the assembly was intelligent," Briscoe says, "the parts knew how to work and fit together. In coordinate systems, the parts would only have a relationship to the coordinate system, not to each other."

The modeling and assembly was actually very easy, taking a total of only eight hours, Briscoe says. "The feature-based characteristics of Pro/ENGINEER and the software's true associativity made it simple." For example, when he wanted a hole he simply used the "hole" command rather than having to create a cylinder and then subtract it from a block. The associativity meant that whenever he changed any part, the whole assembly automatically updated.

Locking on the loading. Before using the Pro/ENGINEER 3-D model, SRAC's Jain used a Pro/ENGINEER interface to draw a 2-D finite-element model of the barrel in SRAC's COSMOS/M FEA software. He knew the pressure loading on the barrel was more of a shock loading. Because this was a linear problem, he chose to follow a modal superposition method of analysis. He modeled the barrel as pipe elements, and the stock as shell elements. "Pipe elements are essentially beam elements with the cross section of a pipe," Jain says. He used the pipe elements because the barrel behaves as if it were a long beam, with most of the response coming from bending.

After establishing the boundary conditions, he ran a 2-D analysis and could see right away that the BOSS altered the frequency of the barrel vibrations.

Frequency and mode shape are important considerations in any dynamic problem. Frequency in loading excites frequency in structure. Jain determined that three-to-five modes of frequency would correspond to the loading in the structure since the time period of loading on the barrel is 0.0014 seconds.

When doing a shock loading, Jain says, it's best to include all modes greater than or equal to the time period of the loading. "Five modes would have contained 80-90% of the response, but I used 10 modes to increase accuracy," he says.

It was now time for 3-D analysis using Briscoe's 3-D Pro/ENGINEER model of both the barrel and the BOSS. After translating the models to COSMOS/M, Jain meshed them with a total of 6,000 tetrahedral elements.

He only used a 2-D model for the stock because the main response was coming from the barrel. "To do a 3-D analysis of the stock would have required 13,000-14,000 tet elements, and there would have been no significant difference in results," Jain says.

Barrel action. The barrel bolts in two places to the stock, so he merged barrel nodes with stock nodes at those locations. Then he did a frequency analysis. COSMOS/M has a mass properties command that calculates sectional properties such as mass, weight, moment of inertia, and volume, among other characteristics. Using that command, he determined that the weight of the rifle without the BOSS was 8 lbs, 6 oz. Actual weight turned out to be 8 lbs. 8 oz. His frequency analysis showed the mode shapes to be smooth, with no irregularities.

Next, Jain put the pressure loadings on the model, applying them where users load the bullets. Then, he let COSMOS/M run a dynamic analysis of the tip displacement without the BOSS, with the BOSS at a zero setting, and with it at maximum setting. Computer time for the analyses took 10 minutes per run. Jain did about three runs for frequency and three for time history. All were on HP 700 and IBM RISC machines.

Without the BOSS, tip deflection maxed at 0.0014 seconds, when the bullet was leaving the rifle. Without the BOSS, tip displacement was 0.32mm in the vertical direction. With the BOSS in the zero position, tip displacement was 0.16mm. In the maximum position, it was 0.12mm. Jain also did acceleration plots and saw that amplitude of acceleration was reduced vs time, which reduced displacements.

Results showed that the BOSS itself makes a difference, and positioning of the BOSS makes a bigger difference.

From a dynamics point of view, putting additional mass on the barrel alters the frequency of the whole system. The peak-to-peak on the acceleration plot lengthens, which reduces displacement.

Setting sights on optimization. "The reason the extra mass works is that it alters the frequency," Jain says. The maximums get shifted slightly. The point where the bullet leaves is not where the tip deflection is at the maximum. The bullet has already left by the time deflection is at maximum.

Now that Browning knows that altering frequency is the key, Chief Designer Rose and his team can improve the BOSS system through computer simulations. Previously, they did the testing in the field.

Adds Browning Chief Engineer Nelson: "Our empirical testing gave us an indication that vibration was the key factor, but we weren't sure precisely what was occuring with the BOSS. Now, software has validated our empirical database, and we can go forward on optimization."