PU to aluminum bonding issues

[Lukecon]

Hi, im currently having issues with the bonding of a cold cast MDI polyurethane, to an aluminum substrate,

i have currently been using the following primers with the results next to them
AD 1147 ( adhesion failure is a hit and miss )
Silbond / AD 1147 ( adhesion failure )
Chemlok 213 ( adhesion failure )
chemlok 219 ( PU to adhesive failure )
chemlok 213 / 219 ( adhesion to substrate is achieved in the 219, but during a peel test the 213 delaminates from the 219 )
i am currently waiting on Cilbond 41 and 48 to trial these primers as they have come recommended for cold cast PU
all PU materials are heated exactly to their processing temps and vac'd to degas before pouring over the substrate, from the sides the PU thickness would be 2mm and ontop of the substrate would be 13mm of PU (thinking a square), during its 15min cycle time the PU elevates to around 80C before maxing around 100C for a further 10 minutes before cooling off ) the PU is then left to cure on a workshop bench for 3 days before demoulding .
i dont currently have access to a oven for prebaking or curing of the product, so what i am trying to achieve is a cohesive failure working on a cold cast.
what im currently finding, mainly when using the Chemlok 219 by itself, is leaving the 219 to dry for ~17 minutes in an ambient temperature of 27c and pouring directly over the substrate, i have a better sucsess rate than leaving it to cure for 1 hr before pouring over the substrate, what i am finding is that when i am pouring in this time zone, i can guarantee a bond between both substate /primer / PU
in the times between 20-30 minutes i can see a film of the PU stuck to the primer, although the bond strength is still able to be removed by hand the PU needs to pass 90degrees ,
im currently looking at the film thickness, as i am only brushing the 219 so the film thickness varies trying to see if this makes a dramatic difference in bond strength particularly on the PU to primer, rather than the Primer to substrate, as I have found the 219 will ensure a bond between the primer and substrate.
Could someone possibly shed some light on why I would be having a successful bond, in these particular time windows, as I am running out of ideas ,

 

[TomDOT]

Has the aluminum been thoroughly degreased? ie - surface contaminants removed.

If so, does it have a nice sharp surface profile? Often something like 50-75 microns. Rounded profiles are less helpful for bonding.

"Sharp" is like: VVVVVV
"Rounded" is like: UUUUUU

Typically the first would be achieved by grit blasting, the second by shot or bead blasting (sharp abrasive versus round abrasive)

 

[Lukecon]

i have grib blasted and also acid etched to ensure a correct surface prep, i also MEK wash all items prior to painting primer on , i can achieve a bond from the primer to substrate but the PU is still pulling away from the primer

 

[TomDOT]

Okay, are you using the primer recommended by the manufacturer of the PU? They appear to be different brand names.

 

[Lukecon]

the manufacture had recommened that i use Conap AD1147 with the silbond as a adhesion promotion, what what i had found regardless of pre work to the steel it has great bond between the PU and primer, but the primer doesnt hold onto the metal substrate, which is why i had gone to using the chemlok 219, as it remains a consistant bond to the substrate, but the PU is a hit and a miss with bonding to the primer

my plan of thought was identify a primer that would hold onto the substate, ( which the 219 is doing a perfect job ) then find out a primer or a secondary coat i can apply ontop of the 219 which will help bond the PU to the primer,

 

[Compositepro]

Aerospace does a lot of aluminum bonding and they get bond strengths to aluminum that are significantly greater than to steel. Aluminum is always surface treated (chromic acid etch, phosphoric acid anodize). Primers are used to protect the surface treatment and do not improve the bond.