Time required on Dynamometer

[Divas]

Hi,
Can someone help me to estimate the time required for the engine tuning through dynamometer for a fixed intake and exhaust setup? This is the first time we are running our engine through a programmable ecu (PE3). Now we want to tune the engine through dynamometer and we need to discuss the amount of time we require on the dyno with the company. Being the first timers we have a little idea about the same and are not aware of the problems that one faces during dyno testing. Can anyone with the prior experience help us with this. Any help will be really appreciated.

[slicktop]

It is going to depend on the skill of the person doing the tuning, and how good of a tune you will be happy with, and how competent the design of your intake is. If you haven't done so already, get it idling decently in the shop. Then get it driving decently in the parking lot. This will do two things: 1. You will gain experience tuning the car. 2. The car will hopefully be fairly reliable when you go to put it on the dyno. Trying to tune an unreliable car can greatly increase the amount of dyno time you will need. For example, you set up some dyno time, spend time transporting the car to the dyno, spend time setting the car up at the dyno, the car overheats in 15 minutes (or, insert other reliability issue here) = a wasted day.

[Racer-X]

An already running car tuned by someone that know's what they're doing might take two hours to get it where you're putting down peak power at WOT in your operating rpm including set up and tear down time. If you have a running car essentially all you have to do is throw the thing on the dyno do a few pulls at WOT, get your spark and fuel close, then go and tune for peak while holding an RPM at several points in your powerband.

If you're trying to tune for anything else or don't know what you're doing count on it taking much longer. Also IMO you can get a decent tune and practice by driving it and then just have to do some final tuning on the dyno. I don't think its worth the time to be tuning anything but full throttle on the dyno, especially not if you're new to it and probably have bigger fish to fry.

Some quick tips:
- Tune your A/F richer than you think you should while testing, then lean it out until you see the power drop.
- Keep the spark on the retarded side so you can advance it until you see the peak and back off to maintain that value.
- When at the dyno add trims to affect the overall trend before tuning cells.

[Jay Lawrence]

To add to the above:

-Use some kind of detonation detection (we used a metal rod bolted to the outside of the head, attached to a tube going into a pair of ear muffs. Worked well)
-You haven't explained who you are or where you're from or what kind of engine you will be tuning, or whether it is an engine dyno or a chassis dyno. That information is obviously important
-If you are first timers, your primary goal is to obtain a safe map, so don't stress too much about power

[Kirk Feldkamp]

Agreed with most of the stuff above. The most important thing for your first time is "keep things simple". Often times, your first time on the dyno with ANY new setup will be an exercise in troubleshooting. You need to ensure that all of your subsystems are working properly before you spend too much time tuning. As Brian points out, it's very smart to spend a bunch of time at your shop before you ever think about going to the dyno working on the starting, idle, and (possibly) some drivability stuff. Personally, I avoid doing too much "street tuning" these days only because it's not very time efficient. I've found I can hit more cells in way less time on a dyno than I can putting around on the street or in a big parking lot. Plus, the end result is generally more consistent (no bumps or hills to get baked into the map), and you get the added bonus of having torque/power data for lower throttle positions (very handy if you're doing more complex vehicle simulations).

First, it makes sense to me to explain what is meant by basic "tuning". The ECU is simply a tool to calibrate when and how much fuel is delivered into the engine, as well as when the ignition events occur. The primary sensor used for fuel tuning is a Lambda sensor (aka O2 sensor). For very basic tuning, simply tuning to target lambda and calling it good can be very effective. The dyno allows you to figure out what the "best" lambda is for any given condition because it will tell you how much power is being generated at any given time. For a rough first cut, you're probably safe tuning everything to a lambda of ~.90. You're not going to be winning any awards for efficiency with that, but you're also not going to blow anything up, or be dumping obscene amounts of fuel out the exhaust. For ignition timing, the dyno's torque readout is one of your few easy indications as to what the engine wants. Tuning ignition without this is generally just guesswork, so read up on "MBT" (minimum best timing) so you understand how to properly develop your timing map. For a rough first cut, there are some very basic guesstimate equations out there, but nothing is going to beat doing it right. At least the equations will give you a rough indication of where to start. The act of tuning a cell requires the engine to either pass through or sit at that operating condition. You will (hopefully) have some sort of logging where you can track throttle position, RPM, and lambda... and be able to go back and scribble down what was happening at various RPM and load conditions. The changes you make to the maps will "correct" the maps based the gap between those observations and your tuning targets. If you don't have any form of logging, good luck, I don't know how that works because I've never done it that way... nor can I see it being done very effectively or efficiently that way!

Since you're new to this, here's my methodology for quickly building maps from scratch:

1. Get it to start and idle. For race cars, using an "idle pocket" on the ignition side of things is a very good idea. Basically, you'll want a "well" of lower advance at the RPM you want to hold the idle. Surrounding that point, you'll want some degree of "ridge" so that (A) if the engine starts to run too low the timing will bump it back up, and (B) as you pull out of the pocket the engine doesn't want to die. This is a bit of a balancing act because too much advance at those low loads can also kill the engine (especially for (B)).

2. Once you have it idling consistently, you can start with other areas of the map. You didn't mention what type of dyno you're going to be using, but there are two basic kinds you're probably going to encounter. One is an inertial dyno, and the other is a variable-load engine dyno (water brake, eddy current, etc). Yes, there are some that combine the two, but let's keep it simple for now. I'll just explain the general method, and you can apply the concepts to whatever you're working with in the real world.

The idea here is that you'll start at lower RPM's, and make sweeps up to higher RPM's. Some dynos allow you to hold a condition steady state, but they will both allow you to do sweeps, so I'll explain how it works with sweeps. Mapping is usually the most efficient if you take a systematic approach at this. I like to work from the bottom up, starting with small throttle openings and progressively going higher and higher as the map gets developed. This keeps things safe move forward because if you're following the plan, you won't find the fueling dramatically off when you finally start testing the higher load areas of the map. I don't know what your throttle pedal setup looks like, but it can be very helpful to at least fashion some sort of temporary, widely adjustable throttle stop so you can hit the loads you want to hit very quickly and easily. It can be very quick to do it "by foot", but the driver on the dyno will need to be able to see the TP% readout so they can close the loop on their foot actuator. :) Generally, once you're to maybe 20%-30% throttle and up, the engine should be able to pull most of the way to redline. Below that the throttle will usually choke, and the RPM's won't be able to climb above a certain point. Just copy that last cell beyond that RPM and move on if that happens. This seems like a good juncture to mention that the TP's you hit should be the ones listed on the load axis of your map. Basically every ECU interpolates between cells, so if you want the tuning to be accurate, you need to take special care to be as close as possible to the "center" of the cells when you are tuning them. This is why I recommend the adjustable throttle stop. This also means it's advisable to have a "buffer" around the load you're tuning, meaning you don't want blank or untuned cells immediately above (or below) the cells you're working on. If I'm working up through the loads, I will usually load the line I just finished up to the next TWO lines... and add about 3% to each. This will keep things safe, and you will be much less likely to have an error based on interpolation to wildly inaccurate cells. Just keep working up to higher loads, and in no time you'll be ready to rock on fuel.

Through this method, I'm usually able to build an entire fuel map from scratch in 4 hours or so (me tuning, someone else driving the car on the dyno). If you have more complex things like turbos, variable valve timing, etc, it can take a LOT longer, depending on how much you want to optimize the setup. I would plan on asking (begging?) for 2 full days for your first time if you think your dyno shop is open to that. This will probably give you enough buffer to do what you need to do, and still have time for the problems that will almost inevitably pop up during the process. Most importantly, have a test plan going in, so your can get the most out of your time on the dyno. You may have noticed the above methodology really only talks about fuel. Once you are done with the fuel, you can go back and begin to optimize the ignition timing. This can be a pretty time intensive process without a variable load dyno, but lucky for you, the ignition timing will generally have less of an influence on the overall drivability of an engine, assuming you don't have something crazy for an ignition map. Focus on getting the fueling mostly correct, and you'll be better off than half of the teams at comp.

Happy tuning!

-Kirk

[Goody]

Thank you for the post[s], they've been helpful for me as well. We'll be starting to tune our F4i here very soon, while we have a stock bike base tune [that it runs on], with it being restricted and on E85, we'll have quite a bit of tuning to do. The different approaches helps me out, luckily our shop is very open to helping us dial it in as well!


As far as the starting timing "well". Performance Electronics [PE3 as well] has starting compensations that I think may cover what Kirk is referring to, for fuel and timing.

[Mbirt]

I believe Kirk was referring to a valley in the spark advance table at the target idle speed and load ranges. I've done this in combination with a lean valley in the fuel map to create a very stable idle without any sort of active idle control. If speed drops or load increases from idle, you take advantage of the torque reserve at the lean and retarded idle with rich and advanced mapping to get back to the calibrated idle. We did this with the clean snowmobile this year and it idled much better than last year's setup with idle state PID's controlling both idle spark and throttle airflow.

[Big Bird]

What a damn pleasure to see some polite, helpful discussion returning to these boards. Kudos to all, it has made my day :)

[Andrew Richards]

Good stuff above but some things that should also be mentioned:
-Try to maintain constant inlet air temp, this is critical for spark timing.
-Try to maintain engine coolant temp at the operating temperature for the same reason as inlet temp.
-Monitor exhaust gas temperatures when tuning your air fuel ratio.
Generally I prefer an engine dyno because these values above can be controlled where as on a chassis dyno you begin to get hot and have to come back to idle to cool down (although this is a good time to make your calibration changes).
Try to spread out your days on the dyno so that you can analyze data, fix problems, make changes and show up with something new to try rather than doing the aforementioned while wasting valuable dyno time. This process can take two hours or two months. Depends on what your goals and capabilities are, everyone has to start somewhere :)

[Kirk Feldkamp]

Matt said it way better than me! Yes, that's what I mean. It works pretty darn well once it's dialed in. It can also be very effective at keeping the engine from doing the classic "FSAE surge" at idle (revving up, falling down, revving up, falling down, rinse, repeat).

Yes, maintaing temps is very nice, but not always possible without some fancy facilities. If you stick to MBT, you'll more than likely have enough buffer when conditions swing in the negative direction (hot). Since you say you're using E85, you've got even more buffer. In my experience with race cars running E85, the fueling is fairly insensitive around your peak power point (more so than gasoline - so you can possibly lean it out a bit more), and really the main benefit is being able to add maybe 3-6 degrees of max advance and pick up about 5% extra power. This is on an already mostly-optimized fuel and ignition map.

EGT's are great, assuming you know how to use them. Unfortunately, there's not a direct correlation to EGT and any specific tuning change you need to make. It's a multivariable system (function of both fuel and timing), and they both have significant influence on your EGT's. Personally, I only tend to look at EGT's to see if something is *way* off. I used it to identify rich misfire conditions before... where the lambda says "LEAN!", but your EGT's are outta sight, meaning there's a ton of unburnt fuel lighting off in the exhaust - so you actually have to reduce the fuel amount significantly to get it back into the right range. Ultimately, the EGT your engine can withstand is a mish mash of component capabilities and engine conditions. I only tend to use EGT sensors on turbo cars where there is a clearly defined max gas temp for the turbine wheel. I've never run into EGT problems on NA engines where the lambdas are in the right ballpark.

I definitely agree on spreading things out. It would be awesome if you could turn things around quickly, but that's not always the case depending on the issue. The hit list of things I always make 100% sure of before I get on the dyno are: water leak checks (0% tolerance for leaks), a properly functioning fuel system (makes target pressure, absolutely no leaks, injectors are bled - usually won't start if they aren't bled properly), solid idle at a reasonably low RPM, and the ability to rev up and drop the throttle without the engine stalling. It goes without saying that you also need to do a "sanity check" on all of your sensors to ensure you have correct calibrations for all of the sensors. Definitely check all your temp and pressure sensors for this. Any error codes should be resolved before the dyno as well. The worst thing you can do on the dyno is spend time dealing with "silly" errors that you could have dealt with at the shop ahead of time. It's a big kick in the shorts if you find out you have a bum sensor once you're already deep into your tuning session and have to go back and redo some or most of your map(s). Dyno time is valuable, especially if you're paying for it!