Thread: Antifreeze Experiment & Testing

Every week someone posts about mixing antifreeze, whats ok to run? can you mix G12 with Peak? with Dexcool? well i set up a little experiment to see what exactly happens when you do.

http://www.peakantifreeze.com/
It says right on the front page "Mix with any color antifreeze". we shall see....

http://www.getahelmet.com/jeeps/maint/dexcool/
some dexcool info. there also happens to be a class action lawsuit against the makers. more about that here.
http://www.dexcoolsettlement.com/




Heres how its going down. i set up two different experiments to deal with two major problems associated with coolants- the "gel" supposedly formed when mixing G12 and anything else, and the galvanic corrosion that takes place over when you put regular old Peak antifreeze in an Audi engine that has many dissimilar metals.

Experiment 1: mixing of coolants.
undecided as how to accurately test this, as gelling occurs over time in the engine when you mix coolants. it has to do with contaminant build up in the cooling system getting trapped in the mixture, and also the heating/cooling cycle and getting pumped thru the radiator and water pump over and over. unfortunately i sold my little 12v D5 pump a few months ago, which would be perfect to circulate the experimental fluid.

i may update this experiment in the future, but right now i have 4 cups-
1- control- 100% G12
2- 50% G12 + 50% Peak
3- 50% G12 + 50% Dexcool
4- 50% Peak + 50% Dexcool

in each of those, i added 50% by volume of distilled water (1 part antifreeze mix to 1 part water)
no real changes have been observed in regards to gelling or changing consistency. when i think of a good way to test viscosity il update this section later.

(its the back row of 4 cups in this picture)





Experiment 2: Galvanic corrosion

Ok so this is a tough one to do on a timely basis. Galvanic corrosion occurs with two dissimilar metals, an electrolytic liquid, and an electric current.

Over time coolant breaks down in an engine and forms substances that act as electrolytes. Couple that with the many different metals our engines are made out of (aluminum head, iron block, various other metals and alloys in sensors and other parts) along with fact that the entire chassis is one big negative electrode (12v ground) and you can see where galvanic corrosion comes into play.

In this test, we are looking to see the rates metals corrode in different soloutions of antifreeze. There are 6 test subjects. (all mixtures are- half 50/50 mix antifreeze + half distilled water.)

1- Control - Distilled water.
2- G12+Distilled water - 50:50
3- Peak (green stuff) + Distilled water - 50:50
4- Decool + Distilled water 50:50
5- G12 + Peak 50:50
6- G12 + Dexcool 50:50
(bottom 6 cups in the above picture, left to right)


To test this in a timely fashion, we need a stronger electric current to speed up any reaction we may see. I dont have 6 months to let the mixtures mix around in my engine. With this setup i got results almost instantly.

I used pieces of aluminum about 1.5" and pieces of mild steel (couldnt find anything around the house that was pure iron, so this will do)


Attached a copper lead wire to each


and for the electric current. 800mA 12v DC transformer


although it said 12v its really 16.4v but it will still work. your cars electrical system is around 14-15 volts when the alternator is running anyways so this is pretty close.



I wired them all in parallel for two reasons- because i dont have 6 identicle power supplies to put on each of them, and by doing it this way, the current will take the path of least resistance and we will get a bigger reaction from the antifreeze with the most electrolytic capabilities.

The mild steel pieces are wired on the negative side of the transformer, the aluminum is the positive.



theres about 15.5 volts sustained across all 6 cups.




Results!:
Ok, well almost immediately there was a reaction in a few of them. here are my notes, i observed the action at 10 minuits, 1 hour, 2 hours, and il leave it going over night and do a final look in the morning.

#1- 100% water-
10 minuets - just about zero action, nothings happening
1 hour - no change
2 hours - few bubbles, not really anything still
24 hrs - still next to zero change, 100% pure water isnt a great conductor/electrolyte
@2hours



#2- 100% G12-
10min -some bubbles on negative side.
1 hour -no change
2 hours -no change
24hrs - very little change, although there is a small amount of green buildup, probably oxidization of the copper, though its only on the part where the copper touched both the liquid and air. possibly getting oxygen from the air to oxidize the wire, not from within the liquid. note- slightly less green buildup then dexcool.
@2hours

@24 hours


#3- 100% peak-
10 min -rapid bubbling on steel, green buildup on aluminum
1 hour -negative side still bubbling, heavy buildup, steel has darkened
2 hours -still progressing same as before, tons of green buildup on aluminum
24 hours - massive amount of pitting in the aluminum. tons of solid buildup in the antifreeze. copper wire was disolved completely. small discoloration in the negative side. test stopped.
@2hours

@24 hours


#4- 100% Dexcool-
10 min -some bubbles on steel, strange milky gel forming around both electrodes
1 hour -same, some green solid forming on the positive side wire
2 hours -same as before but slightly more progression
24 hours - almost the same as the 100% G12 although slightly more green buildup on the copper wire, except below the liquid line not on the surface. il let this go another few days.
@2hours

@24 hours



#5- 50/50 G12+Dexcool-
10 min -some bubbling on steel
1 hour -same, not much change
2 hours -sort of milky around the negative electrode, bubbling
24 hours - roughly no change, im going to let this go another few days and see how it stacks up with the strait G12 and strait dexcool.
@2hours



#6- 50/50 G12+Peak-
10 min -rapid bubbling at the negative side, milky haze around the steel. very heavy green buildup on positive side
1 hour -even more green deposit on the aluminum. lots of action.
2 hours -same as before plus the whole cup turned an opaque milky color. lots of green deposits on aluminum
24 hours - obviously lots of buildup although different from the 100% peak sample. the G12 definately limited the amount of solid buildup more then 50% (concentration of G12:Peak) and the way it built up was much different as you can see in the pics. the corrosion is limited to a small area, as opposed to the 100% Peak where the corrosion/solid formation just spread across the whole dish. slight discoloration (darker) on the negative side, some pitting. test stopped.
@2hours

@24 hours


right now it looks like G12 and Dexcool are very similar in regards to protecion from galvanic corrosion. Peak is garbage as you can see. it just about instantly started corroding. we will see what the electrodes and fluids look like after 24 or possibly 48 hours.

UPDATE DAY 2
Well its confirmed. Peak antifreeze is total crap and has no place in our engines. Although- there was a twist on the results i expected. when comparing the 100% peak to the 50/50 peak and G12, the sample with the G12 in it did a great job at stopping more then 50% of the corrosion that occured in the 100% peak dish. the corrosion was limited to the point of origin and did not spread thru out the disk as happened in the 100% peak sample.
also there was very little pitting on the aluminum in the 50/50 sample, but there was a pink solid layer that formed. i think that the G12 solidified on the metal to stop the reaction of the Peak. Below are some pics comparing those two samples.

il let the other 3+control samples run a few more days to try and get a more pronounced reaction (the two samples with Peak in them corroded so much that it short circuted the whole line to the point that it was only reading ~8v. i cut the wires on those two samples and the voltage returned to ~16v)

100% peak @ 24 hours
Frothy solid formed across the whole dish.



after some cleaning, notice the pitting




50/50 G12 and Peak. notice the localized green stuff and the formation of a pink film like substance and very little pitting.



cleaned up

liquid turned pretty milky, but not nearly as much solids as the 100% peak.

notes- i agree the copper could be taken out of the liquid for a more controlled test, but its important to note that it didnt corrode in all the samples. it had no reaction in the 100% water, and very little in the 100% G12/ 100% dexcool. also there is copper used in our engines, some gaskets and seals. i believe the test is still valid.

and i agree that heating them up to ~185*F would be a bit more accurate, so would experimenting with used fluids with contaminants in them, but there isnt any practical way for me to do either of those. il update in a few days after i let the 100% G12, 100% dexcool, and G12/Dexcool mix and control dish run longer to try and get a clearer difference between them.

UPDATE: its been a week since the test started, not a huge difference between dexcool and the G12, however theres something to note. compare these two pics, first is 100% G12 and the second is 100% dexcool



theres a slight green buildup on the copper in the G12 sample where the copper met the air and liquid. however on the dexcool side, you can see there is some significant buildup on the whole wire all the way around, definately more then the G12. the 50/50 dexcool and G12, was similar to 100% G12. i figure thats a pretty solid result.

now before i declare G12 the champion, i decided to reverse the electrical current, making the mild steel the sacrificial annode (positive side). i figured what the hell, i could either throw everything out at this point or see what happens. il update in a few days with changes. there hasnt really been any change in the piece of steel through out the whole experiment, with the exception of it turning a little dark in a few samples. lets see if it will do something with some positive current thru it.

its also important to note the control- 100% distilled water. its got some white deposits around the steel after 7 days.



FINAL EDIT
Ok folks, there has been some pretty good results with the reversed electrical current. At the end of the test i also measured the Ohms across the dish to determine how easily electricity was allowed to flow.

First off, here is the control dish of distilled water... lots of rusting on the steel.



Here is the 50/50 G12 and Dexcool. Out of the samples i had left, this one had the most corrosion out of all of them, a very thick build up. These pictures are after taking the electrodes out of liquid and letting them dry over night. It has been about 12 days since i reverse the electrical current. Lots of particulate left in the liquid from pieces breaking off of the electrode. This sample had a resistance of 2.7 mega ohms. (Remember the higher the number the more resistance to electrical flow there is, so higher is better in this case)




100% Dexcool. Slightly less corrosion then the 50/50 sample, but still a substantial amount. Lots of particulate left in the liquid and lots of oxidization on the copper. This sample had a 4 mega ohm reading.



And lastly- the G12. It has some light scally build up on the surface, but its not very thick, and no where near the amount in the other samples. There is a little bit of copper oxidization on the part of the wire that was in contact with both air and liquid. This one had the highest ohm reading of 4.8m ohm.




So you may be asking "Well what were the Ohm readings before the test? which fluid resisted breaking down and, when contaminated, allowed more current to pass thru it?"

Well here you go, i had a set of dishes that i set up and mixed as i did these, but just let them sit there without any electricity passing thru them. Here are the readings....



When uncontaminated, both Dexcool and G12 resist electrical flow alsmot equally. However, once you add some contamination to the mix, you can see how much more that resistance dropped in the Dexcool, as compared to the G12. Peak had a crappy resistance value to begin with, so mixing it with any other fluid drastically decreased the over all value of the mix (increasing conductivity, and allowing galvanic corrosion to speed up as well).

That, combined with the amount of decay experienced, i will be using only G12, Audi Approved, coolant in my car. Whatever you do, never mix coolants, as shown this doesnt give you a mix of the two fluids' performance levels, but decreases it even further then if using one fluid on its own.

I think there were some old threads discussing Dexcool and how it is pretty similar to G12 and that it works as long as you flush your system. I have successfully been running Dexcool for over a year now. Its like 4x cheaper!

good work!

This kind of experiment should be used in a physics or chemistry lab in colleges. LOL. Great Job!!!

Good job... But I think your green build up you are getting is from the copper oxidizing... The way you are stating it, is sounding like the aluminum is giving off the green build up...

That is what it seems like, but it's also interesting that it isn't happening in every sample. All I know for certain is that my sisters cooling system looked like hell when I rebuilt her cabriolet's engine for my girlfriend after she ran peak for a year....

Great work. someone has a lot of time in their hands

I only notice it in the ones with Peak... But I cannot tell from others...

Cool experiment, reminds me of AP Chemistry in high school, we had to do stuff almost exactly like that.

This reminds me of a valuable lesson we all learned at a young age; mixing different colors of Play-Doh f*cks everything over and once they're mixed there's no going back. You're stuck with a brownish mass resembling fecal matter while everyone else is making space shuttles and fairytale creature type things during play time. Worse yet, you just dirtied your diaper and have to go to recess with a fat load in your pants. Who ever said life as a first grader was easy?

Note to self: DO NOT mix different colored substances together unless finger painting, then it's ok in moderation.

would it not be useful to also heat the samples to say engine operating temp?

I was thinking the same thing!

Great experiment.

Keith had it right on how to isolate the issue from the copper wire.
The only thing I see possibly being an issue is the heat cycling, I know you're applying voltage to the fluid, but the heat cycling has it's own effects.

This is how productive I would be if I didn't have the internet.

Wicked experiment!

Im curious to see what one kills in the end!

Not to mention, 12V is a bit much... The shop that I worked for, made us test battery ground to coolant... IIRC, it should have never exceeded .5V? I think with 12V, he is just accelerating the process...

I'm curious...

Did you happen to test the voltage at the end of the circuit (furthest from the power source)? The reason why I ask is because I would like to know what the voltage difference is at the end since it looks like you have the G12 & the water furthest away from the power source.

Wondering if that is lowering the voltage enough to make any sort of impact on the results of the experiment.

Ok gotcha. Great test btw!

Glad to see the extra bucks I put towards G12 coolant is paying off