All-Clad G5 -- A Serialized Review

Cooking Discussions Cookware
21

Now you’re being silly. I tested in the same conditions in the same pan, same morning. The pan, hobs, water and room were all at the same ambient temperature and humidity–the same as they are today. They all received equal airflow. So they were/are equal. You want the allergen count as well?

Then you’d quibble over something else, like disproportionate evaporation or the margin of measurement error. I picked 500ml because I knew water holds and conducts heat well, the G5 is half full with that volume (giving the walls some role to play), my graduated cylinder is more repeatable/smaller error margin, etc.

I didn’t wait for any definite time. I waited until the temperature dropped by an arbitrary predetermined number of degrees. I think that does relate to cooking.

From a physics standpoint, it doesn’t matter if you’d tweak those parameters. If you’re angling to state the obvious, let me spare you the trouble: At the instant the gas is cut, the grate is hotter than the induction glass when its power is cut. But that doesn’t answer either the question of the pan’s responsiveness or the role of carryover heat in Pan X’s responsiveness. I can measure and calculate the number of calories of heat added by the grate and glass to a panful of water, but you’d be even more disappointed with the results.

22

Appreciate the lack of sarcasm. Real data can be important… but I wouldn’t compare sports cars based on their towing capacity. I would much prefer to hear about the responsiveness of cookware/cooktops under conditions where it really matters. I know I suggested oil and onions previously, but that is still not an area where responsiveness really matters (like seafood and some sauces).

I don’t have an ideal method for which to test this… much of it is lost on me as I predominantly cook on ceramic, and pans come off the cooktop on to the surrounding concrete countertops.

I do wish you luck with your series, and look forward to seeing what you come up with.

23

The point is, i you want subjective impressions, you’re going to get those along with real data.

This hints at the possible number of variables and permutations. Removing the pan entirely will be better than leaving it on a source of residual heat. And what you set it on affects cooling times as well. I keep a large piece of 1/2" aluminum (a ship scupper cutout) near one stove. I mostly use it as a footed trivet and for thawing foods, but I recently discovered that, if I put a pan of boiling hummingbird nectar atop it, it cools much faster than it does on a glasstop or even a dead coil. The ne plus ultra would be an Anti-Griddle… https://en.wikipedia.org/wiki/Anti-griddle

24

All-Clad G5 – Episode 2

As promised, here is the data for coil, radiant, and for those with a skeptical bent gas with cast iron grates. I’ve also run the same test with removing the pan onto a ½-inch slab of aluminum. Finally, just for fun, I ran it again for a Fissler 28cm Original Profi on induction.

The results are: Gas on cast iron grate: 29:07

Here, I need to explain that the grate on this hob consists of seven equally-spaced, radial “spokes” that point inward with a 2.5-inch opening at the very center. The spokes are 5/16 inches wide and ½ tall. With this 6-inch pan bottom, perfectly centered, each spoke makes surface contact with 0.55 square inches of bottom. So, for 7 spokes, we get about 3.8 square inches of total floor contact (Compare this with >28 square inches of surface contact on a continuous flat surface). The hob itself has 3 concentric rings of jets, the outer two of which burn directly under the spokes; the inner ring burns under the pan only.

Interestingly, the cast iron grate top surfaces never exceeded 220F, immediately after turnoff. Compare with the Ceran on induction, which registered 167F. However, well before the water dropped to 100F, the temperature of the CI grate dipped beneath that of the water, meaning it was no longer contributing carried heat to the pan/water.

Resistive Electric Coil: 32:13

Here, the hob is a 6” round GE calrod coil with 5 full winds and a 1-inch open center. This is a “tighter” wind than comes with modern coil stoves today; the gaps between coils here are narrower than are the coils themselves.

I was surprised to see that, immediately after the power was killed, the coil only registered 280F.

Radiant Electric: 33:25

This is a Frigidaire cooktop with a black, continuous semi-translucent Ceran top. The hob is marked on the glass with a 6-inch painted area, but the actual coil beneath shows itself to be only 5 inches in diameter. There is no fan venting the cooktop, although the case is mostly open, as is the peninsula in which the cooktop is mounted.

I’m glad I ran this hob mode because the Ceran top material is the same as on my induction hob, and since it’s flat and continuous, they both have the same contact surface area.

Aluminum Heat Sink:

For comparison, I decided to do a run a constant 180F and then remove to my aluminum slab at ambient temperature. Here, the heat sink effect was quite pronounced—by the time the pan dropped to 100F, the entire slab was noticeably warm, about 10F past ambient.

In today’s final test, I decided to see how bad it could possibly get with confined layers. I ran the same 500ml of water in my Fissler 28cm Original Profi on induction, to compound the stored heat and insulation value of the glass with envelopment effect of a swaddled disk and nonconductive sidewalls. 37:11, despite a greater water surface area and greater evaporative heat loss.

So, to recap Episodes 1 and 2, the G5 cooling times to drop 80F in situ are, from fastest to slowest:

On aluminum slab heat sink: 18:01
On wire cooling rack: 23:33
On gas (SS grate): 26:14
On induction: 28:29
On gas (cast iron grate): 29:07
On electric coil: 32:13
On radiant: 33:25
Fissler 28cm OP on induction: 37:11

Conclusions So Far

  1. On my hobs, the G5 cooled fastest on one of the gas grates. I attribute this to the tiny surface area in contact with the pan and the resultant exposure to air.
  2. The G5 was slower to cool on induction, by virtue of its full contact with the glass. The combination of heat stored in the glass, full contact and the insulative value of that glass combined to a longer time. A high-volume case exhaust fan might shorten it somewhat.
  3. On a different gas grate made of cast iron, the G5 took a bit more than half a minute longer to lose the 80F. To this I attribute a tradeoff—there was a higher surface temperature, but very little surface contact. The grate itself cooled more quickly than I expected.
  4. Downward response on the electrics was worse, but I was expecting even worse than that.
  5. On my hobs, there wasn’t a clear winner between gas and induction in terms of less carryover heat effect. I would not replace one for the other on this basis.
  6. Short of a crane and trammel arrangement, I can’t imagine a way to completely zero out the effect of stored heat. The circular gas grate comes closest in what I’ve tested.
  7. We will see with future tests, but so far I see nothing in G5’s 1800W/mK conductivity that helped this pan; if anything, the 20 W/mK vertical transfer slowed cooling.

Coming Up: Cooling comparisons with tinned copper.

Request for Evenness Criteria—As I move into trying to assess and compare evenness, it occurs to me that this G5 is so small, any conventional hob is going to make it appear more even than, say, a 28cm or 32 cm version. In other words, this pan is so small, it can’t really be mismatched on the low side. Likewise for my 8-inch copper gratin 6" bottom).

What do the cognoscenti suggest I use for a heat source that would show differences in such a small pan? A fondue or hotel rolltop burner? A tea candle? Something else?

25

Thanks for the testing. I feel I should contribute, in the interest of playing nice and supporting an inclusive community here… :dizzy:

But with a big disclaimer: as a practical home cook, I’m not a major fan of these kinds of theoretical tests.

You start by saying you want to test the pan’s strengths and weaknesses, but then you immediately focus only on responsiveness? The rest is coming?

Imho responsiveness in a frying pan is overrated. Picture this: I’m frying my pork chops in oil and butter, sage and all, splash of wine… :slight_smile: When do I need responsiveness? Most dishes made in a frying pan will be very simple: heat on, put the food in, no lid, cook until done, heat off, take out food. That’s it. Different from a sauce pan, and imho also from a saute pan where you’ll more often have sequencing of cooking techniques.

Related to this: imho one of the key determinants of responsiveness will be the thickness of the aluminum in the pan, as well as the thickness of the ss and g.

Typically, the thinner a pan, the more responsive it becomes - but thinner also usually means less even heating. Your Fissler test already alludes to this: its slow response is a function of the thick aluminum disc.

The trick is to find a pan that has a good balance between the two. If the G5 has say 10% more responsiveness versus a Paderno GG, but only 1/2 of its aluminum thickness, I’d argue the Paderno is the ‘better’ pan.

For drawing conclusions from your tests, it helps to know the thicknesses involved.

2 Likes
26

Bunsen burner, obvs.

As for the Fissler, 37:11 doesn’t seem very long to me, since it weighs 4 lbs vs the G5’s 1.5 lbs. So despite weighing more than twice as much, it only took ~30% longer to cool to the same temperature. Obviously this result suggests that most of the heat is contained in the water itself. Yes, the Fissler has an appreciably larger surface area, but given the mass difference of the pans alone I’d expect a bigger difference in cooling times. I’d also like to see how quickly the Fissler cools on your gas cooktops.

Otherwise, I’d be curious to know what the water temperatures were on all tested cooktops after 5 minutes of cooling. Alternatively, how long did the water take to drop by 10°F on each cooktop?

1 Like
27

Car cigarette lighter.

28

My, you’re impatient. Sorry I didn’t start where you wanted.

As for the importance of responsiveness in a skillet, reasonable minds can differ. The makers, in this case All-Clad, seem to think it’s important, and the concept goes hand in hand with the new material’s extremely high conductivity. A lack of responsiveness is not a good thing per se, without it garnering other benefits.

We may find that G5 punches far above its weight in terms of evenness. We all understand that, for any given thickness, copper is more even than aluminum, so there really is less connection between thickness and evenness than you may imagine. The graphite material is about 4x more conductive than copper, so it’s not outlandish to think that G5 could be as even as your Paderno, yet more responsive. That would leave only heat retention as a potential disadvantage of a thin pan. Hopefully we’ll find out.

Reasonable minds can also differ about “balance”. I believe that resort to valuing the evenness that is (has been?) associated with thick disk enough to prefer its general use isn’t balanced, but that’s just me.

29

The fact that it didn’t do worse shows little more than the effect of evaporative cooling. It was the same volume of water, spread out in the pan version of a wading pool. If someone wants to send me the G5 saute, I’ll compare.

That may be your interest, but I’m not sure what that would show.

I don’t have a bunsen at present. It may come down to a rechaud/fondue alcohol burner or a can of Sterno.

30

For a start, it would show whether the temperature drop was mostly linear or whether there was an initial period with a proportionately slower decline in temperature. Obviously the more data points provided, the better.

Part of the reason for my suggestion to use less water is because I can’t imagine ever adding 1 lb of wet food to an 8.5" skillet. Maybe an 8 oz filet steak, but that’s probably about the maximum.

31

Each thermal “system” would have its own curve, but if you plotted it, I’m 98% sure the lines would not cross. I’ll see if I can redo a couple and time it at the 5-min mark. I take it you want gas and induction?

A big point of using a 500ml half fill was to bring the G5’s clad sidewalls into play. Also, the “wading pool” effect makes immersion probe temperature measurement harder, as I discovered with 500ml in the Fissler 28cm.

1 Like
32

Ideally, yes, thanks. My best guess is that there will be a brief plateau on gas & induction followed by an accelerating decline, followed by a gradual deceleration of cooling off. These data about downward responsiveness in the early stages after the heat is cut may also be more cooking-relevant.

33

OK, but I predict the cooking-relevant part will be a moving target.

34

OK, here’s Episode 3

Prompted by Andrew, I went back and repeated the 500ml cooling test for induction and gas, but recorded temperatures over a shorter timeframe.

Basically, I ran the G5 on induction against the G5 on the cast iron gas grate. I recorded the water temperature every minute for minutes 2-7 after killing the heat. The times and temperatures are:

Induction:
2:00 164F
3:00 158
4:00 153
5:00 148
6:00 144
7:00 141

Gas (Cast Iron)

2:00 168F
3:00 160
4:00 151
5:00 147
6:00 141
7:00 136

Conclusions. #1, The lines did cross. #2, at the suggested 5-minute mark, the G5 on gas had cooled very slightly faster. #3, At least on these hobs with the G5, there’s not a lot of difference in carryover heat

Based on my earlier results, the gas hob with the SS grate would have done better. And I’m sure had the G5 on induction been kept at 180F longer, the glass would have gotten hotter and the system would have cooled more slowly.

Because we’ve been talking about the insulative value of glass and evaporative heat loss, I also ran 500ml of water in the Fissler 28cm OP on indiction, except lidded to slow evaporative loss. When I saw that after 7 minutes, that the water was still at 163F, I could tell this wasn’t a productive use of my time to stand around. So I went grocery shopping, got gas and returned an hour later. Total time to drop from 180F to 100F? 1:33:33, as in 93 minutes, 33 seconds.

2 Likes
35

Interesting stuff @kaleokahu. I just wish the G5 was larger for some evenness testing.

A bit off topic, but it came to my mind. Have you possibly done evenness testing with the Panasonic Met-All for regular 2.5 (2.3) mm bi-metal copper vs eg Demeyere Proline? I would be interested in the result.

I’m not sure if I could expect the regular good bi-metal to beat Proline by much in evenness performance? I remember that Falk copper core was pretty close with Proline in your tests so far, but Proline had it beat in evenness nevertheless as far as I remember. I have a feeling the bi-metal would actually not beat Proline by very much though? Neck to neck even?

I know you like thicker copper generally :).

36

Hi Pertti,

My recollection is that Kaleo tested a bunch of different pans on the Met-All, including the 28 cm Proline, 28 cm Fissler Profi, a 6 mm thick aluminum Stanish omelet pan, and a 3.2 mm tinned copper sauté. Of those, my memory is that the Stanish was the most even-heating, followed by the Fissler, the copper, and finally the Proline in dead last. It was by far the least even-heating of all of these on the Panasonic. I believe his test is archived here: https://forums.egullet.org/topic/157836-panasonic-met-all-induction-hob-a-review/

Considering how much more conductive material there is in the 3.2 mm copper pan vs the 2.3 mm in most bi-metal copper, I’d imagine most bi-metal copper would perform much more similarly to the Proline.

37

Thanks, I could check that, didn’t know it was archived. I just remembered indeed that bimetal wasn’t there, looks like you remember the same.

I also believe the regular bimetal copper would be pretty close to the proline in actual heat spreading ability. Still interested if Kaleo has actually tested it.

I am not planning on buying either here, btw. I am just interested for the sake of it.

38

I believe Andrew’s recollection (of evenness rankings) is generally correct. I did not have bimetal copper to test.

I also believe the tinned copper skillet I tested was a 2mm piece. An extra
.3 mm certainly would improve evenness. As would–indirectly–the SS lining, because less heat would be lost to the environment. Basically, when you use an intrinsically uneven heat source like induction, the highly conductive, thin, empty pan body sheds heat too easily.

One illuminating way to think about these ranked results is balance. Where on the evenness continuum do you want to place your bets? If responsiveness is important to you, then Proline, and especially Fissler, aren’t the best choices. Coer would be better-balanced, thicker copper even more so.

Another is verticality. An induction field makes the heat inside the pan. I believe this aspect, while it is efficient, renders any given pan thermally “thinner” than it would be on a different mode. Put another way, the distance between the food and the heat is shorter. That means the relationship between orthogonal and lateral transfer is altered. This isn’t all bad, but it’s not all good, either

1 Like
39

Correction, the copper piece was a 3.2mm Dehillerin saute. So it had even more surface area than the skillets tested.

Here’s another example of the role of environmental heat loss: I had a 3mm copper comparator made to the same size as a prototype and the Demeyere double plancha. At the long distances involved in the work, the clad Demeyere was more even than the bare copper.

1 Like
40

OK, in Episode 3, we’re on to actually testing Meekah’s little skillet’s responsiveness and evenness.

Even trying to test evenness is a challenge because of a floor space only 6" in diameter. On virtually all hobs, the pan would appear to be very even indeed, since it’s really difficult to find a hob <6". Some Onions suggested a bunsen burner, but what I chose to use were 7 oz. Sterno “Green” cans of alcohol fuel, as used in rechauds and under buffet pans. I chose these because the apertures in the cans are exactly 2" in diameter, and so the test results would be more comparable to a bigger pan on a home hob (replicating an overhang).

I used two firebricks set on edge on which to perch the pan, placed as far apart as they could be and still support the pan–tnere was nothing directly beneath the pan’s center that could skew results. This placement put the top of the fuel cannister 2.25 inches below the pan’s center. Everything was at 70F starting temperature.

Basically, this test was to assess how fast the graphite base of G5 heated both orthogonally and laterally, what the temperature dropoffs were from center to edge-of-floor, and to compare that data with a 2mm tinned copper gratin of almost the same exact size and shape. The gratin’s floor is 6.5" in diameter versus the G5’s 6", so my “edge” readings were taken 1/2" inside the shoulder.

For both pans from a cold start, I measured every 30 seconds up to 2 minutes 30 seconds, then went to whole minute intervals for 3, 4 and 5 minutes. I also did a final reading at 8 minutes, including a temperature reading at the pan rims.

For the G5, the readings were:
Time Center Edge Delta
0:30 150F 109F 41F
1:00 185F 132F 53F
1:30 200F 152F 48F
2:00 215F 177F 38F
2:30 235F 201F 34F
3:00 250F 218F 32F
4:00 296F 244F 42F
5:00 320F 259F 61F
8:00 358F 292F 56F
At 8:00 the G5 rim was 240F, for a Delta T of 118F

For the Havard 2mm tinned gratin:
Time Center Edge Delta
0:30 131F 104F 27F
1:00 152F 125F 27F
1:30 163F 140F 23F
2:00 180F 155F 25F
2:30 192F 167F 25F
3:00 203F 181F 22F
4:00 222F 186F 26F
5:00 235F 210F 25F
8:00 267F 237F 30F
the copper gratin’s rim reached218F at 8 minutes, for a Delta of 49F

Conclusions

I suspected the G5 would heat faster, mostly as a function of its thinness and relative lack of mass. The test bears this out, although the difference was much smaller than I expected. If the goal is to preheat a pan center to 200F for eggs, the G5 got there in 1.5 minutes, versus a little less than 3 for the copper gratin. If the goal is to get the entire floor to a minimum of 200F, then the G5 took 2.5 minutes, the copper a little over 4 minutes.

In terms of evenness, the 2mm copper pan won at every interval. This did surprise me, given that the conductivity of the graphite core is supposedly around 1,800 W/mK, >4x as conductive as copper. The actual rate is undoubtedly lower, given the perforations in the graphite, but even if that rate is only 900 W/mk, I would have still expected the G5 to be more even.

Are any of these results of a magnitude or margin that make a practical difference? Not to me. Maybe there would be qualitative differences if the tests were of 12" diameter pans, but I doubt it.

Next episode will be on how much heat the two pan bodies hold.

3 Likes