Graph B - If we change the scaling and add some grid lines the graph gets even easier to r
The initial runs on the dyno yielded that the engine was producing a very flat power curve from 4,000 rpm to 6,500 rpm. It was also noted that the power was falling off very quickly as the rpm climbed past 6,500. But before we claim a victory we need to look at more data. We made several adjustments to the engine on the dyno trying to maximize the power. Nothing that would make the engine illegal, but we have some room to adjust and change timing, jetting, valve lash, and work with other engine parameters. If we take the horsepower curves from several of the different tests we can see that some of the adjustments affected the power curves in different ways. Once again we will adjust the scaling on the graph and place all the runs into the graph so it becomes even easier to see any differences no matter how small or large.
When viewing all of the power curves arranged in the same graph (Graph C) we could clearly see differences in the shapes of the curves. Remember we are looking at a very small difference in peak power between all the runs. The delta between all the runs was less than 10 hp but the width of the power band or the area under the curves did exhibit some fairly significant differences up to 6 percent.
* HP-2 showed a unique bimodal peak with a hole or a dip in the power from 5,600-5,850 rpm. This run also had the highest peak power over about a 300 rpm range but it also showed a very steep fall off post peak.
* HP-3 showed a reduction in total power and also exhibited the same dip in the power curve over the widest range of the power curve. Although the dip was not as severe it was exhibited over a wider portion of the power curve.
* HP-6 also showed a reduction in power with some significant perturbations at the peak(s) and a nearly sign off at the end of the curve.
* HP-9 curve was very good. The power started earlier in the rpm range and was stronger over a wider rpm range than all of the other curves illustrated. The peak in power was not as strong as HP-2 but it was more stable as the rpm level reached the peak. Early in the lower rpm levels this run showed one to three more horsepower from the lowest rpm to 5,500 rpm and there were no dips from 5,500 to 5,900 rpm.
Graph C - With the data formatted in this way we can see each power curve and get a better
As you perform any power curve analysis you have to keep several things in mind and answer some questions that will effect how you set up and adjust the car.
* What rpm ranges will the car go through on the track?
* Is the current gearing allowing the engine to over rev?
* Is the current gearing too tall and not allowing the engine to rpm in the correct range?
* What will be your lowest rpm and what will the highest rpm be on the track?
* How will the power characteristics of the engine affect how you drive the car?
* How will the power characteristics of the engine affect your chassis setup?
* Springs or torsion bars
* Sway bars
* Is the engine power shocking the setup? (i.e., is the power too violent for the current suspension setup?)Does the driver skill and the power curve match? (This is a very important question to answer, be honest! This is often a difference between a rookie and a seasoned driver.)
* How would you adjust the suspension if the car suddenly had50 more horsepower?
* How would you adjust the car if you suddenly had less power?
Looking at the data and having some conversations with the driver it was discovered that in the car the engine was being spun to over 7,200 rpm at the end of the straight. At that rpm the engine was way past the horsepower peak. Clearly the car was not going to be accelerating as strongly as it was at 6,300 rpm. We need to remember that the faster the car goes the more power is required to just maintain the speed the car has attained to that point. And, the power requirement is not linear it is exponential. The acceleration rate is determined by how much extra power is needed just to maintain the forward motion at the current rate.