Analysis of 2019 NASA Champs Honda Challenge H4 Race with RaceHero

Introduction

Racing is a data-driven sport.  Regardless of the money or car you have, it's possible to make educated decisions on setup changes or tweaks to driving style by using hard work and commitment.  However, without understanding the craft of data analysis, hard work and commitment can often lead to dead ends.

Burning the Midnight Oil with Crew Members Kelley and Andrew at 2019 NASA Championships

In this article, EJ2 Track Rat will show readers how "free" lap analyses from RaceHero can be used to go faster in a race.

RaceHero Logo

How it Works

RaceHero is a free service that shows the provisional lap times, positions, and announcements for races and time trial events with many organizations such as the National Auto Sport Association (NASA) or the Sports Car Club of America (SCCA).

Racers or high performance drivers put a transponder on their car and their racing info is transmitted to this website for the sake of reviewing provisional results.  While the data on RaceHero is not official data used for prizes or points, it is extremely useful since racers can see their own data and and their competitors' data.

Overall Position on Y-Axis, Lap Times on X-Axis

Fast Lap Times

Throughout the weekend, I was keeping an eye on my lap times relative to my competitors'.  I was up against some stiff competition from the West Coast including National Champion, Robert Krider and his teammate, Keith Kramer of Double Nickel Nine, as well as my good friend Carlos Valenzuela and Steve Peterson of Team Beans and Dog Racing.

These Guys Mean Business!

During practice on the first day, with the help of my Crew Chief, Andrew, I made changes to the car after every session.  I did not run my transponder for the first two sessions of five, but I did run it for the third, fourth and fifth sessions.  During the third session, I was about 1.5 seconds off of the leader, Robert Krider with a best lap time of 1:46.882 vs. his 1:45.338.

Results from Session 3 Practice

By the end of the day, Krider had improved by ~0.3 seconds relative to my times.  This suggests my competitor was able to make a significant change to the car's performance or to his driving style which would make my job harder for the weekend.  Regardless, this analysis prepares one to look for more time in more places-- i.e. work harder.

For the first qualifying session, Team Beans and Dog switched drivers from Steve to Carlos and laid down a flyer.  Of note, additionally, was the overall drop in lap times for all competitors, suggesting that the track was "getting faster," providing more grip for drivers.  This is due to cloud cover, ambient temperature, and humidity per experience.

As the day went on, track conditions became further quick as suggested by another overall drop in lap times.  While I wasn't gaining significant ground on the leader for qualifying, I was maintaining a steady deficit utilizing car changes to keep up with a stacked racing team.  In Qualifying #2, again, I was in 4th, but the results were slightly different, running my fastest lap thus far.

Race Analysis

There were two races during the Championship Weekend-- one was a qualifying race and one was the main event for the podium.  

Standing in 3rd on the Podium Against Reigning National Champ team Double Nickel Nine

During a race, the fastest lap time matters, but what matters more?  The sum of the following matters more:

1.) Traffic management

Managing faster or slower, out-of-class cars passing without significant loss in speed/position are facets of good traffic management.

2.) A good start

Avoiding jumping the start which would result in disqualification, having impeccable reaction time when the flag drops, and deciding when to pass or bump help for a good start.

3.) Car/driver consistency

If the driver cannot maintain their level of performance throughout a race due to a lack of cardio or mental conditioning, they will fall off in speed.  If the car is not set up to make the tires "happy," the car will fall off in speed.  A combination of the two is key.

4.) Flag management

You can be leading a 45-minute race for the entire time.  But if you pass just once under a yellow, you will be disqualified.

5.) Defense

How to stop a faster in-class car from getting around you.

6.) Offense

How to pass an in-class car from behind and make it stick.

RaceHero can be used to measure all of these things-- whether one at a time, or collectively.  For example, a good start would be suggested in part by a good lap time for the first lap relative to competitors or a large gain in position.  Car/driver consistency can be gauged by a drop or lack thereof in lap times over the course of the race.  A defending driver may maintain position in front of an in-class car with a faster best lap time.

Photo of #ChelseaTheCivic Leaving the Keyhole at Mid-O (Photo by Herb Lopez of @DriveNASA

For race analysis, I typically take all of the RaceHero data I can get for my competitors and put it into an Excel Sheet like a true engineer would.  Below you will see that I copied all lap times from RaceHero, converted them to total seconds to make it easier for Excel to analyze, then I measured standard deviation of hot, green laps (not first or last lap...no standing yellow laps) and put it in a table:

Getting my Data Nerd on

For the first lap, it's clear that I had a very good start.  I decided to bump Kramer whose start was not as good to keep Steve on my inside while Ken had the best start of all of us, getting the jump on Keith, Steve, and myself to second place.  

However, leading into China Beach, Ken was passed back by both Keith and I with Steve still in tow of Ken.  The rest of the first lap was a game of traffic management.  All of these factors lead me to the second-fastest first lap time.  However, with nearly a 20-HP deficit to my competitor Krider, a better time would have been difficult to achieve:

Starting Lap Times

Next, we'll look at the beginning laps of the race.  In the beginning laps of the race, it's important to play offense, attempting passes on your competitors while the tires are still fresh.  It's also important to get by slower out-of-class traffic as quickly as possible.  During this portion of the race, heart rate and adrenal release should be highest.

Photo by Herb Lopez of @DriveNASA

Below you'll see a graph of overall lap time in seconds vs lap # for Steve, P4 and myself, P3.  For the first half of the race I laid down a killer start and close to the fastest lap I could for my second lap.  After that, variation in lap time was caused by out-of-class traffic.  However my good start allowed me to maintain a comfortable lead:

Analysis of Race Start

In the middle of the race, I try to conserve my tires as best as possible so I can lay down flyers at the end of the race.  This is also clearly noticeable in the next graph I'll show below.  What's important to remember is while you're saving your tires, your competitor in the spot behind you may be catching up.  This was exactly the case as Steve, in his lightweight CRX began to close the gap:

Mid-Race Analysis

By the end of the race, one may note that Steve and I were running very close times.  Keen eyes will also observe that Steve's fastest overall lap time was somewhere near 1 second faster than my fastest overall lap time, suggesting this driver/car combo had more speed than I.  This makes sense because by the end of the race I was driving as defensively as I could within racing rules.  It's important to avoid blocking or other foul play when employing proper defense, but it's also important to maintain position, obviously.

End-of-Race Analysis

Driver and car consistency among the pack will lastly be analyzed.  To do this, we look at standard deviation and average lap time per competitor.  Any laps with yellow flags, red flags, restarts, and aborted starts are thrown out.  Additionally, we do not consider starting and finishing lap from RaceHero when calculating these statistics.

Since this is averaged over the entire race, this heavily favors analysis of driver consistency over car consistency.  Car consistency is better analyzed in segments as shown above.  However, car consistency is a variable feeding into this analysis.

Overall Race and Consistency Analysis

As shown above, Steve and I had the best consistency over the course of the race (+/- 0.80 and +/- 0.69, respectively).  We had an awesome, clean battle that ended up with me in P3 and Steve in P4.  The race video is here:

Champ Race Highlights

Thanks for the read, all.  Looking forward to my next blog post.

#ChelseaTheCivic, Part 3: Aftermarket Headers and Hondas

Last season, I was in a rush to get my car together.  I skimmed over a lot of details while doing so in an effort to pass competition school with the National Auto Sport Association (NASA) and to race with my friends in the Honda Challenge H4 Series which we were reviving in the Northeast.

In the process, I ordered a DC Sports header with a 2.25" collector for an exhaust system that had a diameter of 2.5".  I was struggling to make it to my first event of the season at Lime Rock Park to test the car.  And on my final weekend before trailering up to Connecticut, I was faced with the dilemma of making this header work with my exhaust system.

With no flux wire or gas, and minimal welding experience, I welded the monstrosity you'll see above.  This monstrosity is a 2.5" PLM (Private Label Manufacturing) test pipe.  I thought I had a 2.5" collector on my header, but this is when I first discovered that I was mistaken.  As a result, the collector on my test pipe did not fit and I was forced to improvise.  So, I cut the flange off of my 25-year-old catalytic converter, cut the flange off of my brand new, shiny test pipe, and welded it onto the test pipe.

As you may be able to see, this last-minute concoction left a leading edge entering into the test pipe, providing for turbulence and most likely a loss of power.  On the dyno at Lime Rock Park after a race, my car had less horsepower than the three other H4 cars dyno'd-- most likely due to this and my janky intake, which I will not discuss in this post.  Additionally the welds were not very good as you can see below...however, the test pipe held up for a full season of racing.

When the off-season came, I knew I had to put this on my list of improvements for the 2019 racing season.  So, I started by purchasing a $211 PLM header off eBay.  

If you'd like to see more photos like this, you can follow me on Instagram:

This header is a replica of an SMSP Tri-Y Header.  It has a 2.5" collector and is a 4-2-1 which leads to less turbulence and higher flow than the 4-1 DC-Sports header I was previously using.  It's also fully stainless steel, very light, and tig-welded together.

The reason I was able to get this header for so cheap is because it is an open-box sale.  PLM and 1320 Performance are two brands which both sell a lot of headers for Honda.  And they have sales like this going on a lot of the time.  

Think of it.  

Your Facebook friend car meet fanatic wants to make 5 extra horsepower on the butt dyno.  He or she goes on eBay and buys a brand new, shiny, tig-welded header that has no damage at all and intends to put it in a car that is likely daily driven, and if not, is street-driven in his or her leisure.

This header is shipped through FedEx, UPS, USPS, or another service that throws the box around, sets it on fire, or dips it in a bath of acid (they damage it).  It gets to the customer, and they no longer want it.

As a racer, you can take advantage of this scenario by buying the scratched or minorly dented header second-hand, and you have a new, but unboxed item delivered to your door for a fraction of the cost.

Next, I purchased a new PLM test pipe.  This extendable test pipe was ~40-50 dollars and its length can be adjusted to fit your new header to your existing exhaust.

I mounted the header, mocked up the test pipe and got to work cutting and welding.  Having a vise is great for almost anything in the garage.  I even use it to hold pipes or other objects that need welding as I don't have a welding table, I have a wooden table.  The picture below shows one section of my three-piece test pipe in a vise with a white mark around it where it is to be cut.

  After cutting, I then mocked the test pipe up between the header and the exhaust one more time and welded the pipe in the appropriate spot.  This time I used flux core wire, to preclude accelerated rusting as seen on my older test pipe.  A better weld with less slag and less porosity would likely be achieved with gas, but this blog is about doing racing on a budget, so forget about it.

Mounted up, this is what the test pipe looked like.  It fits well, not pulling the exhaust too far forward or pushing it too back.

I also noticed that the SMSP tri-y header was touching the oil pan after I installed it.  Even in a 25-40 minute race, conductive heat transfer from a header to an oil pan could be bad.  You see, exhaust gases leaving an internal combustion engine's cylinder are typically around 1,000 degrees F (based on passing knowledge).  I knew I had to increase the clearance between the header and the oil pan.

Oil pans are cheap and made of soft, malleable, durable steel.  Headers are expensive and made of more expensive stainless steel.  When push comes to shove, one must give-- and in this scenario, I made the pan give with a hammer-- I mean I massaged it gently and carefully as I am a skilled fabricator:

In the above picture, the horizontal lines at the bottom of the pan mark where the header and oil pan were originally touching.  The dots at the top of the oil pan tell me in which order I need to torque the oil pan bolts.  5 bolts tells me that a given bolt is the fifth bolt in the sequence, 17 dots tells me it's the seventeenth bolt.  There are over 17 bolts on a B Series oil pan.

The hardest work was complete, but the last thing to do was to install the O2 sensor.  The problem with this is that on a 1992-1995 Honda Civic with a B Series Engine from an Integra or something similar is that the O2 sensor wire for the B Series Engine is too short for the Honda Civic engine wiring harness.  As a result, you have to lengthen it.  I began by cutting my brand new Denso sensor in half.

Next, I grabbed some wire, measured from the sensor to the connector in the engine bay, and cut the appropriate amount to connect the O2 sensor on the Integra engine to the Honda Civic harness.

Here's a tip: when you take apart any electronics for anything, save the wires.  Good wire is hard to come by at local hardware stores and if you can find it, you have to pay for it.  I save wire from cars that I part out, or machines that break around the house.  I then take these wires and use them in automotive projects.  I haven't spent money on wire in three years.

Then I grabbed my soldering gun and got to work.  I cut all wires needing connection, stripped all said wires of their insulation, soldered all wires together, and added heat shrink around the soldered connections to prevent damage via water intrusion or other external factors.

Here is the heat shrink I applied.  You can get it from a local hardware store, a local auto parts store, or online:

Lastly, I taped the wires together and wrapped the lower part of the wires that sit next to the header with insulation.  I was careful to wrap the soldered joints with tape so that they would not flex.  I was also careful to ensure an equal spacing of tape about the length of this loom of wire.

The finished product is shown below:

Lastly, I installed the O2 sensor...

...And I fired up the car.  The car was difficult to start at first, but ended up idling and running with no check engine light at temperature for several minutes.  The car was sitting for months not having turned over and it didn't even need a jump start.  I love this Honda.

I have some actions going forward to make the car faster, but for now I am focusing on getting it on track next month (April of 2019) at Lime Rock Park for testing with NASA NE.

I appreciate you all reading and wish you the best.  Peace!

B Series Axles-- Problems and Solutions

EJ2 Track Rat: B Series Axles

If you road race or do HPDE with a Honda that has a B Series motor such as a B18B1, B18C5, B16A2, etc, then you’ve likely run into axle problems.

Figure 1: Integra Axle with Torn Outer CV Boot

Many cars with continuous velocity (CV) axles, even street cars, encounter torn boots every once in a while. But when you throw in hard braking from 100 MPH to 60 MPH once every minute-and-a-half, or when you subject your CV joints to excessive stress with a lowered car, or when you toss R comps and excessive lateral loads at your vehicle, you’ll quickly find the limits of physics.

Figure 2: Damien’s Dirt-Nasty Low “EK” Hatch Race Car

Excessive heat from rotors, loads from cornering, and aggressive articulation angles all result from the conditions mentioned above. With the added heat and stress, the components inside of the CV joint typically wear in an accelerated manner-- and for those who are racing, that could mean a DNF or last place. Either way you’re losing points toward your championship and wasting money with a broken car on track.

This writeup provides a solution for those of you who hate your axles. Do you want to drive on R comps? Do you want to be dirt-nasty low? Do you want to crack rotors with heavy braking? Well, you’ve come to the right place.

Figure 3: The Chaos That is Racing

Problem 1: Grease

The first problem associated with many parts-store brand axles that are made for B-Series-equipped Hondas is that they likely don’t have enough grease in them. Even if they do have enough grease, one has to question how high quality said grease is. When a single CV axle costs $35, do you trust it to hold up to the stresses of racing or track days?

Figure 4: These SurTrack Automotive Axles are Cheap at ~$35 from RockAuto

We solve this problem by replacing the old, crappy grease originally present in your axles with high-quality grease. For this DIY, I chose to use Redline CV-2.

Figure 5: Redline CV-2 “High Moly Content” Grease

Redline CV-2 is the grease of choice for rebuilding these troublesome Honda B-Series axles as it is made by a reputable company, withstands temperatures up to 500 degrees Fahrenheit, high pressures, which are common in CV axles, and because it contains an “organic moly” as stated by Redline. This organic moly is likely a compound referred to as molybdenum disulfide. Due to its molecular structure, it provides superior lubrication like graphite but it also adheres to metal surfaces very well.

Figure 6: My Friend Alex Lubing up an Axle with Redline CV-2

Problem 2: The Heat

As I mentioned at the beginning of this writeup, the axles are subject to a lot of stress on a FWD race car. Heat is transferred from hot rotors to the axle spindle through the hub and heat is generated by the bearings in the CV joints as they rotate and articulate. The effects of this heat can be exacerbated by other factors mentioned in this article such as poor grease and poor manufacturing tolerances in cheaply made axles.

If the grease can’t handle the heat, its lubricating properties degrade, subjecting the bearings in the CV joints to excess friction-- compounding the problem. If the ball bearings are pitted, the bearing races are cracked, or the cages are cracked, the axle may begin to vibrate and eventually fail catastrophically.

Figure 7: Cracked Cage on a Subaru CV Joint

To solve this problem, we firstly use better quality grease. However, we can also vent the CV joints.

Venting CV joints has not been empirically tested to the best of my knowledge-- however, it is a practice that has been employed in the Honda road racing community for a long time by reputable axle manufacturers such as Gator, RAxles, Insane Shafts, and Driveshaft Shop. While the efficacy can be debated for days, it still doesn’t hurt to understand the theory and construction of axle vents.

Figure 8: Venting the Outer Boot of an Acura Integra Axle

When the axle is rotating at a high speed, it is believed that the centrifugal force generated by this behavior disperses any excess grease around the inner diameter of the axle boot. This, in theory, creates an air space in the center of the CV joint. Heat from the hot, rotating CV joint is transferred to the air space, and if the air is not vented, it is believed that this can accelerate axle wear.

Figure 9: Diagram Showing Air Pocket at Center of CV Boot

To evacuate this hot air, it is believed that a small “vent” (which is actually a tube inserted between the boot and axle shaft) can be used.

Figure 10: Diagram Showing Added “Small Tube” Which Serves as a Vent

Problem 3: Manufacturing

I’ve taken apart brand new parts store axles that have failed on my race car and discovered that they were adequately greased, but failed anyway. This, and other anecdotal evidence I’ve read online and heard about from other B Series drivers leads me to believe that “they just don’t make ‘em like they used to.”

This last section isn’t me saying that a parts store axle with high quality grease can’t last. Contrarily, I have seen and heard of people regreasing parts store axles with high quality grease and having great success at a reduced cost. However there is a generally true inverse relationship between quality and cost with auto parts that suggests machining tolerances and material quality may be compromised for the sake of pricing-- especially with economy-priced axles.

So, let’s not focus on how true the above statements are. Let’s just make sure that our axles are well made. The easiest way to do this is with visual inspection and with precision measurement tools.

Figure 11: A Starrett 0-1” Micrometer I Bought for $10 on OfferUp

When disassembling your axles to regrease and vent, measure any components you can and visually inspect the friction surfaces. Visual inspection should always be implemented as it is cheap and easy if you know what to look for. If you don’t have precision measurement tools, it’s not the end of the world, but they do help.

Starting with the outer joint of an Acura Integra CV axle, it consists of six ball bearings, an inner race, a cage, and an outer race which is integrated into the spindle. This type of CV joint is called a “Birfield joint.”

Figure 12: Diagram of Outer CV Joint

Disassemble the outer joint and begin visually inspecting the components. The following video is a great DIY on how to disassemble the outboard joint of a CV axle: https://www.youtube.com/watch?v=3-R11jtnyV8. Things to look for during inspection include: metal pitting, hairline fractures and cracks, and discoloration.

Figure 13: Discoloration on the Spindle Caused by Overheating

If you can see pitting on the metal surfaces, it could be due to contaminants in the grease such as sand that entered through a torn boot or corrosion from moisture. If you see discoloration, it is likely that grease was not lubricating effectively in the affected area and the metal is now distorted. If you see cracking, the affected area may have overheated at some point or may have been constructed from cheap metal.

Figure 14: Ball Bearings with Spalling

A well-manufactured CV axle, such as an OEM unit, will show minimal signs of wear on the friction surfaces. I have disassembled OEM Acura Integra Axles with over 150,000 miles on them and torn boots with nearly perfect looking wear surfaces.

Figure 15: Outer Races of Said 150,000 Mile Axle with Minimal Wear

Once you’re finished your visual inspection, measure whatever you can with precision measurement tools. Engine builders use this process when disassembling motors. Critical dimensions are measured to the nearest one-thousandth, sometimes the nearest ten or one-hundred-thousandth of an inch to gauge the health of various components. If the components are found out of spec, they are sent to a machine shop where they can be brought back into spec with specialized machinery.

Figure 16: Using an Outer Diameter Micrometer to Measure Ball Bearing Diameter

For the outboard joint of a CV axle, one easy component to “blueprint” is a ball bearing. Take several diameter measurements of the ball bearings and compare the results against all bearings in the axle set. Mark each ball bearing with dots so that outliers can be identified. Record your results for future rebuilds.

My OEM axles that I rebuilt all had healthy looking ball bearings with outer diameters that were the same to one-thousandth of an inch.

Figure 17: Recording Ball Bearing Diameter Results

The process is the same for the inboard CV Joint. Disassemble, clean, visually inspect, and blueprint what you can. The inboard CV joint used on B Series Hondas and Acuras is known as a “Tripod joint.” The picture below shows a cutaway view of the side of the inboard joint on the left, and on the right, a cutaway view of the front.

Figure 18: Tripod Type CV Joint (Image from: http://what-when-how.com/automobile/universal-joints-automobile/)

Below you’ll see the inboard spindle that houses the majority of components comprising the inboard joint assembly. Again, this is an OEM unit and looks nearly immaculate after years of abuse.

Figure 19: OEM Honda Tripod Joint Housing (“Cylindrical Pot Chamber”)

When I disassembled the inboard joints, I only blueprinted the outer diameter of the “semispherical rollers” as outlined in Figure 18. But I could have also disassembled these “rollers” and blueprinted their internals as well.

Figure 20: Semispherical Rollers Cleaned and Recently Inspected

Again, when blueprinting the inboard side, the more measurements you can take the better-- but remember, all of the measurements are relative (unless you’re an engineer at Honda and have access to the dimensions you need!) Regardless of what the measurements are, record them and keep them for future axle rebuilds. More data is always better.

The Finished Product

Disassembling, cleaning, inspecting, greasing, venting, and rebuilding your old axles can be a hassle, but if you’re looking to save a buck, the tips in this writeup should help. When you buy a project car, save the OEM axles! When your friend wants to throw out their OEM axles, save the OEM axles! And if you’re at a junkyard and you see OEM axles, save the OEM axles!

#SaveTheOEMAxles

Figure 21: An Assembled, Regreased, Vented, and Blueprinted OEM Honda Axle

Thanks for reading all. Your time is greatly appreciated.

Figure 22: Chelsea the Civic

Helpful Links

CV Joint Types: http://what-when-how.com/automobile/universal-joints-automobile/

How to Disassemble a Birfield Joint: https://www.youtube.com/watch?v=3-R11jtnyV8

How to Replace CV Boot: https://www.youtube.com/watch?v=-S-ofMoWsr0&t=329s

Band Tool Operation: https://www.youtube.com/watch?v=mBsqyH6xhw0

How Does Moly Grease Work: http://www.nitronine.com/how-does-moly-grease-work/