Properly testing full undertray on 2011 Ford Ranger

[freeside]

Folks,
I'm currently fabricating a full undertray / belly pan for my 2011 Ford Ranger extended cab, 2.3L pickup truck. I already have the front end piece completely shaped and am working on the finishing the bolted connection details. The first piece goes from the front fender to about 5-1/2" behind the lower front control arms and is 41" long. The truck has an overall length of 16'-11.6", so there's still a lot of work to do. I'm using 3mm Alupanel as the prototype material but am considering moving to Versitex as the final version. Versitex is a thermoplastic composite material with MUCH higher strength and impact resistance, in addition to being lighter and potentially easier to machine, although Alupanel is incredibly easy to machine.

My point in starting this thread, is to start gathering information on how to perform credible A-B-A testing on something so complicated to install/uninstall.
In the next week or so I'll be working up a written test protocol in combination with a QC checklist as a start. This will be my first A-B-A test and I want to remove as much error as possible by writing it all down and having a step by step process so I don't have a brain-fart halfway through the process and miss something. On that note, is there not an archive of test forms, written test protocols etc, shared somewhere on this site? I'm used to having a server full of documents, forms, templates etc at my fingertips at work, but it seems that this is a major component missing from forums...

Current issues:
1. Between A and B tests, it will take me approximately 45 minutes to 75 minutes to install the entire belly pan. Should I leave the truck running in order to maintain temperatures and cruise control set point?
2. Uninstall between B and final A test. Same issue although I'll need less time to pull it off than I did to install it.
3. I don't know about course length yet, but I figured 15-20 miles each way for 30-40 roundtrip miles for each section of the test. Say 45min to 1hr for each roundtrip test, 1 hour for install, and 1 hour for removal, I'm looking at a total test period of 5 hours. This doesn't include time for taking all the measurements before each segment of the test, or time to drive back to my house to do the install/uninstall. Now I'm looking at 8 hours total if things go without a hitch. It seems like weather, traffic and temperature variations are going to be too drastic across this long of a time period for reliable test data. Suggestions?

I'm shooting for final development/vetting of test protocol and the completion of the full belly pan around 1 month from now.

In the next couple of weeks, I'll have a blog up to host documents, photos etc and will try to keep this thread updated.
Thanks in advance for your feedback.
-Stephen

[aardvarcus]

I don’t know if you already have or not, but just be sure you think about and monitor the heat from your exhaust as it relates to heating up the underside of your car. You don’t want to create an oven that has your gas tank inside. I won’t beat this to death again here, but search for heat and bellypan if you want more information; there are several good threads on the topic.

As for testing, A-B-A on a reliable and repeatable route same day is the best. Look at some of the tire size tests for some good examples.

Though not as accurate, week long tests on the same route (i.e. the exact same route commuting to work) with one week one way and the other week the other way can also give reasonable accuracy. The week length helps to average out small discrepancies as long as the weather is reasonably similar and you throw out any outliers. The advantage here is you can record the data from your usual travels without having to do dedicated drives just for testing.

[JRMichler]

Your largest source of error is temperature and wind. I chose A-B-C-D-E... testing for my mods. The results were averaged over a full season, partly because I am more interested in total real world results than highway MPG at a particular speed.

My summary thread is: http://ecomodder.com/forum/showthrea...yon-17070.html

[MetroMPG]

Originally Posted by freeside

My point in starting this thread, is to start gathering information on how to perform credible A-B-A testing on something so complicated to install/uninstall.

Agreed -- it's difficult to test mods that are difficult to re/re. I only A-B tested my undertray, and it was only a cardboard mockup. (Easy to pull off fast, but not re-attach.)

Is your goal to measure the MPG change (if any), or just to know whether there's an improvement?

Because if you only want to know whether A is different than B, you could use coastdown testing instead. In a fraction of the time you're prepared to spend on highway runs you could get a lot of shorter coast-downs in for comparison.

Should I leave the truck running in order to maintain temperatures and cruise control set point?

You shouldn't have to keep it running to preserve the cruise set-point. (At least I don't have to on mine, but I use a kill switch to stop the engine. Not sure if the set point would stay if I went from RUN to ACC.)

Also, it's not just consistent engine temp that's important - transmission oil, tires, bearing temps all factor in -- the whole drivetrain. So it's important that whatever you do, there's an equal amount of driving/warm up time (assuming steady ambient temp) before each set of runs.

[freeside]

Originally Posted by MetroMPG

Is your goal to measure the MPG change (if any), or just to know whether there's an improvement?

MetroMPG: Thanks for the good feedback. As it's generally agreed upon that belly pans produce improvements in MPG, this isn't really in question. I'm looking to accurately quantify the actual percentage gains of this belly pan on this truck. In doing research on this matter, I'm having a hard time finding this kind of data, so I figured I would develop it myself. My oberservation is that part of the reason there is a lack of information on belly pan improvements of MPGs on trucks is that:
1. it's complicated/difficult to fabricate an entire belly pan for a truck
2. it's complicated/difficult to accurately measure the improvements

As the belly pan seems to have the largest potential MPG impact on trucks aside from the aero-topper, I was surprised to find such lack of data.

As I'm new here and don't yet have a reputation, I don't want to appear rude in saying this, but here goes. I want to provide the community the most reliable and undisputable testing results possible. I'm not interested in less accurate methods, or less acceptable methods. From my research, A-B-A holds the most clout with the aero-modding community outside of wind tunnel/chassis dynomometer testing. Therefore, I'm asking the community for help on minimizing the questionably of a quite complex test, to reduce the potential skepticism once results are provided. Does that makes sense? I don't want to go to all this effort to later have everyone throw out my testing results.

-Stephen

[UltArc]

From your last post, you want to know the easiest way to get reliable results.

Well, make sure you have data recording/tracking- SG/UGii, so you can calculate the distance. A-B-A testing works best. So, I would advise finding a nice route you can use cruise control and just let it go. Try to make sure it is not a busy time, and won't be interrupted by cars on your rear end or slowing you down.

Two things you can think about. Like Metro said, as things warm up, they get better. Engine, tires, tranny, the whole shibang. So you can do A, then B, then A2, and focus more on the change between B and A2, as your changes should be minimal from nature (wind, temp, so on), and you A2 should be more so efficient in general from warming up more.

Example:
Run A netted 20 mpg
Run B netted 25 mpg
Run A2 netted 23 mpg

Focus B vs A2, B may not be optimal, but it is a better reflection than A vs B.

Another idea is to warm the vehicle up very well before hand. Drive 15 or so minutes away, then back, and do your test. The Insight is at lean burn temperature within a mile at 35* F, but the transmission and tires are definitely not. LEt your vehicle sit for as long as it takes to equip/remove the gear, don't hop right into the test.

Then your ABA should be more spot on.

And, theres always ABABA, on the same route or technique. The biggest point to show credibility is actual proof, such as video, and data. For reliable data, your best benefit is to reduce variables. It's near impossible to eliminate them, but if you do your best to record and eliminate, the better your results.

Edit:

Note this is the internet, and many of us have no proven credibility. Other than the three people I met for my fuel economy challenge, I have no proof. Heck, my tank wasn't sealed when I left, I could have put some fuel in along the route, even. I didn't, and I don't think the other two did, but we have little proof without being there. The bigger/more arrogant the claim, the less likely we are to believe, but with data, most things are reasonably taken. We all know underbodies work...you just want to quantify for your vehicle. Post up your results, and we'll all be happy to work with you

[smokey442]

Simply form a good data base with your existing configuration. The larger the # of samples the more accurate results. Compare that to your post modification results. Don't waste time going back to A when you already know what A is.

[UltArc]

Originally Posted by smokey442

Simply form a good data base with your existing configuration. The larger the # of samples the more accurate results. Compare that to your post modification results. Don't waste time going back to A when you already know what A is.

Not going back to A leaves the opportunity to miss things. I am not trying to argue, I just want to make sure it's clear that information can be missed by not going back.

[MetroMPG]

Going back to A is definitely preferable if you want to have results that will stand up to a skeptic.

It confirms the validity of your "base" measurements, so you know any fuel economy difference you may see in the B runs is a result of the thing being tested and not some other factor.

[aardvarcus]

Going back to "A" helps to average out any slight changes in the weather pattern or road conditions. For example if it takes 30 minutes to do a route and 30 minutes to change your configuration, it is possible that the temperature has warmed up (morning) or cooled off (evening) in the hour between the tests. Assuming these changes are close enough to linear, the ABA will factor most of that out once you average A1 and A2.

Make your route long enough to be credible. Also, since you are testing aerodynamics with the belly pan, a higher speed test will show the differences more. Interstate or highway is best. Also, minimize things you have no control of. Try to avoid traffic, red lights, etcetera. Testing at an odd hour on a four lane would be best, minimal traffic to avoid and an extra lane to avoid them in. Also, use cruise control to the extent possible if you have it. Cruise should react basically the same on all three runs, to prevent you from accidentally accelerating hard up a hill on one run and easy on the next.

Lastly, as others have stated make sure your car is fully heated up before you start. Using a gauge, monitor your intake temperatures as well as your engine temperatures. Due to heat soak, my intake temperature continues to rise long after my engine temperature has flattened out.

Edit-
Oh, and one other item that would make your belly pan test really great. Get a non-contact thermometer and after each run get the temperature of the Engine Pan, Transmission Pan, Gas Tank, and Rear Differential.