All-Clad G5 -- A Serialized Review

Meekah was kind enough to donate her 8" All-Clad G5 (“G” for graphite, “5” for the SS-A-G-A-SS layup) for testing. What I hope will follow in this thread is a piecemeal evaluation of the pan’s strengths and weaknesses, features and foibles.

In this Episode 1, I ran some tests intended to assess All-Clad’s brags about how tremendously responsive G5 is. In my parsings of that particular brag, I conclude A-C bases it on the intrinsic, astonishingly high thermal conductivity of the very thin center layer of graphite in the pan’s core (probaby around 1,800W/mK). I also wanted to test and quantify the effect of “carryover heat” from various hobs on downward responsiveness.

Specifically, in this first responsiveness test, I wanted to see how fast G5 cools on its own, and then compare that with the times when the pan is left on deactivated hobs after the heat is cut. Yesterday I did that first run (just taken of the heat to air cool), plus on induction and gas hobs. In Episode 2, I’ll do the same for open resistive coil and radiant.

I set the pan containing 500ml of water on the hobs to attain a constant 180F, on the verge of simmer, on induction, and then gas, hobs. The glass and grate were as hot as they were going to get. Then I killed the power/gas without moving the pan, and timed how long it took for that 500ml to shed 80F in open air.

The G5 was first removed from the hob and placed on a wire cooling rack. The time to drop to 100F was then measured to be 23:33.

On induction, the pan’s water dropped back to 100F in 28:29.

On gas, it dipped to 100F in 26:14.

One conclusion here is that even this (allegedly) “lightning fast” G5 pan isn’t very downwardly responsive with ½ liter of water or moist food in it, no matter what you cook on.
This is true even if you remove the pan altogether. I think swaddling anything in SS is going to dramatically slow downward response.

Another conclusion is that there is enough heat stored in flat glass in full contact with the pan (along with its insulation value) to slow the cooling considerably. On my vented induction hob, the G5 held the heat by about 5 minutes longer than air cooling.

A third conclusion is that my gas grate didn’t hold as much heat, or slow the G5’s cooling, by as much as I expected it would. I attribute this to a couple factors. First, this particular grate is stainless steel, and so would not conduct heat back to the pan as fast as would have cast iron. Second, it’s round in cross section, meaning there is only a minuscule area of actual contact between it and the pan bottom (probably less than with the close-mesh cooling rack). Third and conversely, there was ample air circulation. Whatever heat the grate added slowed cooling by less than 3 minutes compared with no added heat at all. When I get time, I’ll repeat the test using a hob with a cast iron grate, but that grate is also open-center with minimal surface contact with pans, especially one this small. Obviously, grates that are massive and/or make a lot of surface contact would likely conduct enough more heat to lengthen cooling times.

I also think that if the pan was preheated very fast using maximum heat, the glass and the grates (maybe not the coils) would not have stored as much heat. For example, remember the parlor trick the appliance stores used when induction appliances first came to market? They’d take a skillet, sawn in half, and fast-fry an egg in it, while an ice cube sat unmelted on the glass a few inches away. It’s not that the glass doesn’t get hot, it’s that the glass hadn’t had time to be heated by the pan. In this test, the glass over the induction hob attained 167F, and after 5 minutes without power, was still at 118F.

Coil and radiant data to follow…


Hi Kaleo, for a basis to understand the time it takes to drop from 180F to 100F, have you done this with other pans (regular clad SS, cast iron, whatever) and if so how long do they take to lose heat under similar ambient temp/humidity conditions?

I’ve never tried anything like this before so the timescales, 23 or 28 minutes, don’t register well in my head.

1 Like

I’ve done similar tests, but not exactly like this one, because Meeka’s G5 is so small. I only have one other pan of the same size and dimensions–a 2mm tinned copper gratin. I was planning on working that into the mix for evenness, but now that you mention it, I’ll run this same test on it just for more perspective. Do you have a prediction?

I chose half a liter of water because it tends to show differences in larger scale, is repeatable, and you can vary the test to calculate the amount of heat imparted by the glass, grate, etc…

1 Like

Is this a skillet? If so, who would put a half liter of liquid in it? Maybe a better test would be with a Tbsp or two of oil (something we would actually heat in a skillet)?

I’m waiting for the autopsy :carpentry_saw::hammer_and_pick::bomb:

Oh, no - I have not a clue. I was just wondering if there were baseline experiments and what their times might be. I might test on my stove tonight (resting on-hob (gas) and on-cooling rack) if my wife relents on her “Honey-Dammit-Do!” list (I feel like I deserve a bit of time off after my fence gate rebuilds, anyway).

My home RH is right at 50% (based on tested wet bulb temp - I had to fix my AC a few weeks ago) and is normally 73°F. RH is probably one of the biggest factors affecting time to cool.

1 Like

Hi Scott - I think the half-litre fluid is basically a proxy for a pound of food, of whatever sort.

I chose water because most foods are moist, and water’s known properties make for easy energy calculations. You could use oil, provided its thermal properties were known, and you measured the volume extremely precisely.

Or you could try scrambled eggs or pancakes and drive yourself crazy…

But it is not how I (or most of us IMHO) cook in an 8" skillet, so maybe run your test again with real food that would be used in that pan. I really don’t need/want to know about a bunch of water in a small skillet, but would be interested in a saute/caramelization in butter or oil with the proper quantity of food.

The focus on water is misplaced. It’s the pan’s thermal performance that’s at issue. Water is an ideal medium for testing that.

Fear not, though, I will be cooking real food in the pan, too.

Feel free to do your own testing.

Water holds a lot of heat… and way more than something like sautéed veggies (or even meat). If you think it takes 20-30 minutes for those kinds of things to cool off, I am dubious about your conclusions.

1 Like

It seems as though you are also testing (in a roundabout way), the heat carryover properties of an induction cooktop vs. gas grates (albeit only YOUR gas grates) vs open air, which is actually quite interesting.

That’s not it, as I understand it. It’s just that a glob of veggies and meat is not reproducible from one glob of veggies and meat to the next glob of veggies and meat - one will have more water content, the other less, one more fat content, etc.

So using a known amount of water is a good proxy for getting relative information pan-to-pan. I don’t read the OP as saying that X minutes for water cooling equals X minutes (or any minutes) for food cooling.

That said, as I alluded to above one would need consistent room temp and humidity (and airflow, which I failed to mention) to be able to compare pan A to pan B etc.

1 Like

Yes, that’s an incidental advantage to assessing downward responsiveness on the 4 different hobs. I just started with the two most obvious modes. Remember, A-C touts the pan as having lightning fast responsiveness. Let’s see how G5’s downward responsiveness compares with the Cu 8".

I think that’s it. The glass is a poor heat conductor. So most of the heat from pan was lost in other direction, where as your stainless steel and cast iron grate will provide plenty more surface to lose that heat.

Hi Kaleo,

It’s not at all clear what you’re trying to accomplish in this experiment, especially in the absence of key data points omitted from your discussion.

As CCE pointed out, it would be essential to know the room temperature, relative humidity, and airflow (e.g., extractor fan, open windows, fan, A/C unit, etc) during each of the three successive tests to make accurate comparisons. It would also be helpful to know whether you conducted them in the same part of the kitchen, as airflows and temperature can vary from spot to spot. Also, what was the order of the tests – did gas come first, followed by induction, etc? I imagine the room’s humidity, and possibly temperature, would have increased with time.

There are other data points you do not report, such as how long the pan spent on each burner before the heat was cut. These would be helpful to know, as well.

I’m also perplexed by your experimental design. I agree that water is a good baseline medium to use to test heat retention, but you’ve filled an 8.5" skillet with over a pound of water. My 7.5" All-Clad is more than 2/3rds full of water with that much added; I imagine the G5 was looking almost similarly full. This leaves very little of the sidewalls exposed to air. As the pan was filled with more water (by mass) than the likely mass of the pan itself (minus handle), you’ve basically confirmed the high specific heat of water. If you instead wished to test the responsiveness, or evenness, of the pan, you would need a much different test.

Further, did you track the temperature at specific timepoints? It would be much more helpful to know the temperature at pre-planned timepoints than the time to a specific, and rather arbitrary, temperature target (100°F). If you did track the temperature decline along the way, what did you discover? A more cooking-relevant test might be to reduce the water by 1/2 and track the temperature decline during the first five minutes after the heat is cut, recording the temperature at 30 second intervals.

Finally, If you really wanted to control for other variables to test the G5’s responsiveness & evenness, you would need to test it against an All-Clad D3 or Copper Core of similar diameter. Testing against a copper pan of smaller diameter will tell you very little indeed.

1 Like

You’re right… I’d just be more comfortable with tests that are closer to what one would actually do with the pan. Maybe weighed amounts of EVOO and chopped white onion (with each test cut from similar areas of same white onion). EVVO (as well as most organics) has a pretty high emissivity, making it easy to test the surface with an IR Therm at various time points.

While this may not be as technically accurate as 500ml of water, I think it would better indicate the responsiveness of the pan/cooktop combo.

I did track those environmental things, but they’re hardly “essential”, because they were all the same. I also tracked the heat loss over time. They’re, like your critiques, are not relevant.

It’s completely patent what I was trying to accomplish: test how downwardly responsive Meekah’s G5 was, with contents in situ on various hobs. We know now that there was slower cooling on induction than on gas. Today I’ll do radiant and coil.

I think one aspect of G5 is already emerging in my results: the graphite is an excellent lateral conductor, but a very poor orthagonal one. That is, it MAY be even, but it transfers heat in and out about as well as brick or glass.

You apparently missed or didn’t retain “Serialized” and “Episode 1”, so there is more to come.

I don’t know what to say about your comfort in a non-sarcastic way, except to note that I am actually cooking food in the pan. I’ll report on that, but don’t expect real data.

All due respect, but “same” ≠ equal. By way of extreme example, consider the possibility that you set up a high-powered fan adjacent to the cooktop. In that case, convective cooling of the pan on the gas cooktop & cooling rack would have far exceeded that on the induction cooktop. Any significant airflow would likely advantage the former two, in any case.

How do you account for the temperature and humidity of your kitchen remaining constant throughout? In any event, why not report these data and let us come to our own conclusions about their significance? Withholding them simply invites skepticism.

If you really wanted to devise a test of downward responsiveness that would be relevant to cooking, I’d suggest starting with a smaller quantity of water and a shorter period of time. Waiting for half an hour to declare a significant difference smacks of designing an experiment to suit an end.