WMI dyno mapping

Hello everyone, and especially Andre, thank you for the work you do and for this forum. Since I couldn't find an exact guide on WMI tuning anywhere on the internet, I decided to study the subject and methods used by people from various sources. Below is a compiled procedure that should serve as a sort of guide for a tuner encountering WMI tuning for the first time or who has been doing it by trial and error so far. I would like to ask the experienced tuners present here to comment on the procedure from a technical perspective and clarify some inconsistencies—see the inconsistency mentioned right in the introduction. I believe this thread will be useful for everyone considering this modification.

Contradictions / Divergent Logics

Order of steps – Fuel → Boost → Ignition timing, vs. Fuel → Ignition timing → Boost

Procedure approach: First, fine-tune the AFR with WMI active, then increase boost pressure (while monitoring knock and AFR), and only "once you safely reach the target boost values" proceed to advancing ignition timing.

Dyno mapping approach: Step 3 (Fuel) → Step 4 (Ignition timing, based on the measured IAT drop) → Step 5 (Boost, only once steps 3 and 4 are already stable) – meaning boost is increased only after fine-tuning the ignition timing in a given cell, not before it.

_____________________________________________

Step 0 – Preparation, before touching the map at all

Set/verify the IAT sensor location downstream of the injection nozzle (not upstream of it). This is a prerequisite for every subsequent timing-related step.Measure the actual methanol % in the tank (hydrometer/refractometer) and record it — without this, none of the later calculations are meaningful.

On the Aquamist, manually force 100% flow and measure the actual nozzle output (ml/min) at operating pump pressure. This figure is the input for all fuel calculations, not the catalog/datasheet value.

Have the failsafe resolved and tested. One of the failsafe types according to the vehicle's capabilities (e.g., control unit input, connection to the compressor bypass, etc.).

Step 1 – Configuring the trigger on the Aquamist

Action: set the trigger to a combination of boost + injector duty cycle (IDC), not pure boost.

Reason: boost alone doesn't tell you how loaded the engine actually is (e.g., climbing a hill at low boost but high IDC). The combination lets you keep boost as the primary trigger (a cheap, fast signal) while IDC acts as a backup/supplement that adds dosing even in situations where a pure boost trigger wouldn't respond.Set the onset point (start of injection) at roughly 1/3 of your planned maximum boost, and the full-flow (max) point well before peak boost is reached — ideally at or before the torque peak.

Monitor: don't measure anything power-related yet; just confirm the solenoid/pump actually switches at the correct point (indicator light, flow meter, listening for it).Conclude: if injection comes on late relative to boost onset, move the onset point lower. This gets dialed in first, since every subsequent step depends on it.

Step 2 – Zero point: verifying the map WITHOUT WMI influence at low boost

Action: drive at low boost, below the onset threshold, where WMI isn't spraying yet.

Monitor: lambda, ignition timing, IDC — these must match your baseline (no-WMI) map.Conclude: if they don't match, something is wrong outside of WMI (in the baseline map) — this must be fixed before proceeding. Never "fix" the baseline map by adding/removing fuel to compensate for a WMI effect.

Step 3 – Fuel: calculate, don't dial in by trial and error

Action: for every boost cell where WMI is spraying, calculate the chemical fuel-equivalent being added (measured nozzle flow × current WMI pump duty cycle × methanol %). Subtract exactly this equivalent from the existing fuel value in the map — not "a bit" followed by watching the wideband.

Monitor: lambda/AFR after this calculation — it should shift only minimally from target, since you subtracted exactly what WMI actually added.

Conclude:

If AFR comes out significantly richer than expected → either actual nozzle flow is higher than you measured (re-measure), or the mixture is richer in methanol than you entered (re-check the specific gravity).If AFR comes out leaner → the same in reverse, or the onset/ramp is set differently than you think (go back to Step 1).The wideband/lambda sensor here is only a check on the calculation, not a tool you use to hunt for the correct fuel value by trial and error.

Step 4 – Timing: based on actual charge temperature, not a flat addition

Action: log post-injection IAT alone, as a function of boost/RPM, with WMI active. From this log, derive how much the charge is actually being cooled compared to the no-WMI condition at the same boost.Based on that measured temperature delta (not an estimate), add timing to the corresponding cell — a larger IAT drop means a larger safe timing margin, a smaller drop means less margin.

Monitor: knock feedback correction (FBKC/FLKC, knock count), not just the wideband.

Conclude:

No knock and stable FBKC → add another degree of timing to that cell on the next pass.As soon as correction/knock appears → back off one degree and hold that as the ceiling for that cell. This is your "maximum" for that boost/mixture combination — not a number from a forum thread or a friend's tune.Work cell by cell (divide the map into boost/RPM segments), not just at a single wide-open-throttle point.

Step 5 – Boost: add only once Steps 3 and 4 are stable

Action: increase target boost by a small increment (1–2 psi), ONLY in the region where Step 3 (fuel) and Step 4 (timing) are already done and knock-free.

Monitor: IDC again (so injector duty cycle safety limits aren't exceeded), knock, post-injection IAT, and whether the WMI ramp (onset/max on the Aquamist) still matches the new boost range.

Conclude:

If the new boost pushes the curve peak past the WMI full-flow point sooner than intended → move the Aquamist's max point up slightly, so full flow is available at the new peak boost as well.Return to Steps 3 and 4 for the newly added boost range (recalculate fuel, re-find the timing ceiling) — this is an iterative loop, not a one-time step.Repeat Steps 3→4→5 in small increments until you hit one of your safe limits (IDC, EGT, turbo/engine mechanical margin, or your target power).

Step 6 – Validate the whole map, not just the peak

Action: run through the entire RPM and load range (on the dyno and/or on the road/via logging), not just peak load.

Monitor: the transition zones around the WMI onset point (where the fuel/injection ratio changes fastest) — this is where holes or timing dips most often show up, exactly as happened in that thread with Rodwrx23.

Conclude: if there's instability in the transition zone (AFR spike, timing dip), refine the resolution of the Aquamist's table/ramp curve in that region — don't "fix" it by adding fuel elsewhere.

Step 7 – Fail-safe test at the end, not as an afterthought

Action: with the finished "maximum" map, artificially disable/disconnect WMI (unplug the nozzle/solenoid) and run through low-to-medium boost.

Monitor: what the IAT-based timing table does, and whether the engine stays in a safe state (no knock, no extreme leaning-out).

Conclude: if the test reveals risk (e.g., insufficient timing retard when WMI fails), either reduce your target maximum boost/map aggressiveness, or implement an appropriate fail-safe — and this is the definitive power ceiling you can run on this configuration (turbo, fuel, mixture).

The resulting "maximum-power map" is therefore not a number you dial in by matching the wideband — it's the point where your safety margin runs out in one of Steps 4–7 (knock, IDC, IAT margin, or the fail-safe test) before the turbo's or engine's physical limits do. That's where you stop, back off one step, and that's your final map.