8YSZ via F4-MBX porosity issue

[firegod]

Dear Colleagues,

Need an advice on 8YSZ deposition using F4-MBX plasma torch. My spec requires roughness of 12-16%, I tried to acquire it using the following parameters:
Argon 38 nlpm
Hydro 9.5 nlpm
Current 550 amp
Carrier 5.0 nlpm
Feed rate 70 gpm (Metco 233C -125+45)
Torch travel speed 400-500 mmps
Double injector feed from 2 separate powder feeders

Examples of reference/acceptable structure are on pictures 1 and 2.

I tried to replicate it with the parameters stated above with 2 alterations subject to my system's layout:
1) instead of 2 feeders I used one with T-connector
2) reduced carrier gas from 5 to 4 nlpm to merge the standard double spraying spot into one long vertical spot (to optimise the programming)

As a result I got structure shown on Picture 3. Comparing to the first two ref pictures it contains a lot of disperse porous phase (marked on the pic) which increases the coating porosity significantly (up to 24-26%) although most of parameters are the same.

Next I reduced the feed to 60 gpm and increased Hydrogen supply to 11 nlpm. It helped me to reduce the undesired phase quantity but overall porosity didn't go below 20% (see pic 4).

My questions are:
1) What is the marked phase in the coating? Is it 233C unmelts or can it be just dust getting caught onto substrate in between passes? I so far evaded using air nozzles on the gun, so to say, to reduce part programming efforts.
2) Can different porosity results acquired on same spraying parameters be caused by different hose/cable lengths on my system and reference systems?
3) Was it a good or bad idea to merge 2 injectors stream by reducing carrier gas?
4) Which way must i go with the parameters to further optimise the structure and acquire the desired 12-16%?

Images:
#1: https://imgshare.io/image/UZWj6
#2: https://imgshare.io/image/UZjKc
#3: https://imgshare.io/image/UZ8Fe
#4: https://imgshare.io/image/UZPay

Thanks!

[Spraypowderengineer]

I would not use a T tube. You never know how much powder goes either way. Better to use two injectors with two feeders.

Feed gas per injector controls how deep you inject. Deeper is more melting, so better DE but lower porosity.

Use 8 mm nozzle, not 6mm (in case you do).

If you do not find a process window, vary electric current and start from new.

This powder has a weakness, it has low mechanical stability, generates fines and creates pulsation of real feed rate. Such short time overload of the plume can be one of the serveral reaons for the high ununiformity of the degree of melting in your coatings. Try another powder, if the problem persists.

Hope this helps you