The Temperature Stall

[infernooo]

To those who are big on food science during meat cookery, I thought the following discussion on why internal temperatures stall during smoking / low and slow barbecuing was quite interesting... anyone have any thoughts or words to add? Criticisms?

http://www.bbq-brethren.com/forum/showthread.php?t=90123

[infernooo]

To those who are big on food science during meat cookery, I thought the following discussion on why internal temperatures stall during smoking / low and slow barbecuing was quite interesting... anyone have any thoughts or words to add? Criticisms?

http://www.bbq-brethren.com/forum/showthread.php?t=90123

No-one? ;-)

It does contain some misinformation and some offhand comments, but later in the discussion they do get more in depth and post some interesting finds from research:

[jackal10]

I think the stall is more likely due to the fats melting than collagen to dissolving to gelatine.

Nathanm would know the definitive answer, since he is also a prize winning BBQ pit master

[slkinsey]

As far as I know, the conversion of collagen to gelatin is endothermic. So, if the amount of thermal energy you are putting into the meat is small enough and there is sufficient collagen present, there would be a plateau in the temperature of the meat while the reaction happens. Pork fats melt at around 40C +/-5, and of course this would take fusion energy, but if that were to cause a "stall" we would then expect to see it at around 40C. But of course, melting pork fat and rendering pork fat out of adipose tissue are not the same thing. Adipose tissue, meanwhile, is made up of fat and collagen and so rendering fat out of the adipose tissue will be dependent in part upon breaking down the collagen. The observed temperature of the "stall" at 74C, which is much higher than the melting temperature of pork fat,makes me believe that the "stall" does not meaningfully result from melting fat.

Edited August 17, 2010 by slkinsey (log)

[infernooo]

I think the stall is more likely due to the fats melting than collagen to dissolving to gelatine.

Nathanm would know the definitive answer, since he is also a prize winning BBQ pit master

As far as I know, the conversion of collagen to gelatin is endothermic. So, if the amount of thermal energy you are putting into the meat is small enough and there is sufficient collagen present, there would be a plateau in the temperature of the meat while the reaction happens. Pork fats melt at around 40C +/-5, and of course this would take fusion energy, but if that were to cause a "stall" we would then expect to see it at around 40C. But of course, melting pork fat and rendering pork fat out of adipose tissue are not the same thing. Adipose tissue, meanwhile, is made up of fat and collagen and so rendering fat out of the adipose tissue will be dependent in part upon breaking down the collagen. The observed temperature of the "stall" at 74C, which is much higher than the melting temperature of pork fat,makes me believe that the "stall" does not meaningfully result from melting fat.

Thanks for the replies, I would be very interested to see what Nathanm has to say (perhaps Douglas as well if he gets the chance!)

[nathanm]

I think the stall is more likely due to the fats melting than collagen to dissolving to gelatine.

Nathanm would know the definitive answer, since he is also a prize winning BBQ pit master

Thanks for the vote of confidence! We have done studies of meat in a differential scanning calorimeter (DSC). This is a device that looks for "stalls" as you heat things. It puts constant heat in, then watches how the temperture increases.

The most dramatic sort of stall is a phase change (ice melting, water boiling) but a DSC is made to have great precision in how it looks for ANY difference in heat versus temperature.

When you have a chemical reaction - like collagen converting into gelatin - it will show up with a DSC.

We have measured this transition and it is real, as is fat melting. Both contribute. However, neither one of them is big enough to really account for the "stall" observed by BBQers.

I think that a better explanation is wet bulb temperature. Basically, if you have a wet thermometer, it reads a lower temp than a dry one, due to evaporation (unless relative humidity is 100%).

Food mostly cooks at wet bulb temperature - until the outside is totally dry. I suspect that is that is the origin of the stall.

[nathanm]

As far as I know, the conversion of collagen to gelatin is endothermic. So, if the amount of thermal energy you are putting into the meat is small enough and there is sufficient collagen present, there would be a plateau in the temperature of the meat while the reaction happens. Pork fats melt at around 40C +/-5, and of course this would take fusion energy, but if that were to cause a "stall" we would then expect to see it at around 40C. But of course, melting pork fat and rendering pork fat out of adipose tissue are not the same thing. Adipose tissue, meanwhile, is made up of fat and collagen and so rendering fat out of the adipose tissue will be dependent in part upon breaking down the collagen. The observed temperature of the "stall" at 74C, which is much higher than the melting temperature of pork fat,makes me believe that the "stall" does not meaningfully result from melting fat.

I basically agree, but the issue is complicated.

Pure fat (i.e. rendered lard or suet) does have a relatively low melting point. It is not a sharp number because it is a mixture of different fats as pointed out in other posts.

When we talk about melting fat during cooking, we really mean rendering, which is largely about converting the collagen in the fatty tissue. Raw fat as it occurs in piece of pork or beef is NOT pure fat - it has lots of connective tissue in it, which is mostly collagen. So rendering fat out of a piece of meats is about both melting the fat, and breaking down the collagen. This is all pretty obvious if you try to render a piece of beef or pork fat - it takes a long time and a lot of heat. Compare that to melting some (already rendered) lard or suet.

The conversion of collagen into gelatin occurs across a wide range of temperatures, at a rate that increases exponentialyl with temperature. That is why we must cook 48 hours at 55C/130F to get tenderness that we might get in an hour at much higher temperature.

Anyway, it does take some heat to melt fat, and it does take some heat to convert gelatin. That will cause SOME stall, but I don't think that is large enough to be a real explanation.

My guess is that in a typical smoker cooking a big pork shoulder or brisket, the wet bulb temperature stays below 74C until such time as the crust of the meat is very dry, at which point dry bulb is more important. This is, of course, discussed at length in my book. This seems like a much better explanation to me.

[PedroG]

Hi Nathan,thank you for your profound explanations.

If your wet bulb / dry bulb hypothesis is correct, this temperature stall would happen in a BBQ oven, but not if you cooked an identical cut of meat sous vide (or braising btw) at a water temperature higher than 74°C. Did you do the experiment?

If memory serves me, collagen is enzymatically converted to gelatin by collagenase, which gets inacivated at about 60°C; a slow process which takes one to several days depending on age-related cross-linking of the collegan. At or above 60°C collagen fibers which have not yet been enzymatically converted to gelatin will swell and shorten (a process reversible by cooling or stretching of the fibers) and squeeze liquid out of the muscle fibers. At about 74°C collagen undergoes a physical phase transition (like melting ice) which is much faster than the enzymatic conversion to collagen, and which would explain the temperature plateau.

The book Food Proteins by Fox and Condon (page 279) says: Slow cooking at a prescribed temperature is often advocated as a means of tenderising. This can be done at either of two temperature ranges, at 50°-60°C which favours the ageing effect or at 80°-100°C where the interstitial collagen is destroyed. These two tenderising temperature ranges are well separated by a sharp minimum temperature of tenderisation at 66°C to 68°C. To hold meat in this temperature region would be the worst possible treatment because it favours collaeen shrinkage and an increase in toughness.

Please correct me if I am completely wrong.

[slkinsey]

I think that a better explanation [of the BBQ "temperature stall" is wet bulb temperature. Basically, if you have a wet thermometer, it reads a lower temp than a dry one, due to evaporation (unless relative humidity is 100%).

Food mostly cooks at wet bulb temperature - until the outside is totally dry. I suspect that is that is the origin of the stall.

Very interesting! I'm not quite sure I get it, though. As I understand it, the temperature of the pork shoulder or brisket or whatever is measured with a needle probe stuck deep into the center of the meat. How would this ever not be "wet bulb" until the meat was completely desicated?

Pure fat (i.e. rendered lard or suet) does have a relatively low melting point. It is not a sharp number because it is a mixture of different fats as pointed out in other posts.

When we talk about melting fat during cooking, we really mean rendering, which is largely about converting the collagen in the fatty tissue. Raw fat as it occurs in piece of pork or beef is NOT pure fat - it has lots of connective tissue in it, which is mostly collagen. So rendering fat out of a piece of meats is about both melting the fat, and breaking down the collagen. This is all pretty obvious if you try to render a piece of beef or pork fat - it takes a long time and a lot of heat. Compare that to melting some (already rendered) lard or suet.

Right. I guess that was more or less my point: that melting fat out of a piece of meat is really more about breaking down the collagen than it is about melting the fat.

[nathanm]

Here is more detail, replying to Pedro and slkinsey.

Heat gets into food from the surface. If the surface is wet, it will be at the wet bulb temperature. This depends on two things - the food staying wet near the surface, and also the relative humidity in the air around the food. If the humidity is 100%, as it is inside a SV bag then wet bulb = dry bulb. If the food is underwater, then there is no air to worry about and the temperature is again the same.

The surface is wet because there is water in the meat, but over time the surface layers dry out. This tends to raise the temperature until with a very dry crust on the meat very little evaporation will occur and the temperature will be at the dry bulb temperature.

The reason this affects the interior temperature is that the interior can't be any hotter than the surface. So if the surface is stuck at say 74C, then the interior can't get any hotter. As the surface dries out, the surface temperature rises, and sometime later the core temperature will too.

As Pedro points out this would not happen inside a SV bag (or in a combi oven or CVAP oven in low temperature steam mode.) Yes we have done the experiment (sort of) - we have recorded temperature in meat cooked SV or in combi for long intervals and we do not see a stall.

I say sort of because if you really wanted to check this out you would need to take two identical pieces of meat (say brisket), treat them indentically (dry rub and so forth) and then cook on in a smoker. We haven't cooked a whole brisket side by side. We did smaller pieces of meat, and we didn't do perfect comparisons. So a skeptic could still.

Most smokers / BBQs have no humidity control, and people don't know what wet bulb temperature is. We show in my book to rig up a wet bulb thermometer. We think that is how you should control your BBQ, and once you do that we think that reports of the "stall" will disappear.

I have several smokers and we have wet bulb temperature in them and you don't see a stall.

We have also done experiments using a convection oven, or a very fancy smoker, where you can show the wet bulb effect easily. Take a convection oven (or combi oven in convection mode), set the temperature to say 60C/140F and put meat in, it plateaus at the wet bulb temperature.

Finally, if you put meat in a DSC you can see a bit of a stall due to the various chemical changes that are happening but it is NOT a single fixed plateau, nor is the total amount of heat big enough to cause the reported hours-long "stall" that BBQers see.

Conversion of collagen into gelatin is complicated it does NOT occur at a single temperature - it is a reaction that occurs at an increasing rate over a range of temperatures, all the down to just above the animal's original body temperature. It does not occur "at" a certain temperature.

Sorry but I can't resist saying that all of this is covered in depth in my upcoming cookbook.

Edited August 18, 2010 by nathanm (log)

[slkinsey]

Very interesting. So let me see if I get this right...

For a given piece of meat in a barbecue-style smoker the temperature of the meat will slowly rise until it reaches a kind of "temporary equilibrium temperature" where evaporating moisture maintains the surface of the meat at a more or less constant temperature via the "wet bulb effect" (the inside of the meat also plateaus since peak internal temperature is dependent on surface temperature). The temperature of this temporary plateau should be somewhat different depending upon the ambient temperature and humidity in the smoker, but I imagine these variables are similar enough in barbecue-style cooking to account for the broad similarity observed in this temperature plateau. So the meat sits at the plateau temperature until the surface of the meat dries out sufficiently so that evaporation is no longer able to cool the surface of the meat effectively in the ambient environment (this is now "dry bulb"), at which point the surface temperature and consequently the internal temperature can rise. The time that the meat spends at the "temporary equilibrium temperature" until the surface dries out is the observed "temperature stall."

That is not only incredibly interesting, but also a fun debunking of what everyone thought was a revelatory explanation of the science behind this observed effect! I wonder the extent to which the "temporary equilibrium temperature" could be altered by introducing moisture into the smoking chamber.

Oh, and don't worry. I am most certainly buying the book!

[nathanm]

Yes, you got it right.

I am going to run a brisket today or tomorrow to get the definitive test.

[slkinsey]

I wonder how this might be affected by the "wet mop" technique of barbecuing, where the surface of the meat is periodically wetted throughout the smoking process. Might this be a way of maintaining an overall lower internal temperature? It seems like you might be able to keep the wet bulb effect going for a long time this way... although I imagine that there are limits to its effectiveness (thin liquids won't adhere to the surface very well and what does adhere will evaporate off quickly, whereas thicker liquids may not evaporate at a high enough rate to provide sufficient local cooling).

[nathanm]

Yes, wet mops matter. Having water in the smoker (as is a water smoker) matters. Having the smoker be full of meat raises humidity, so that matters.

Humidity is the source of a lot of the variation that we see in cooking processes.

Professional smokers used for meat processing control humidity for this reason. Combi ovens and CVAP ovens do too.

[e_monster]

I wonder if this is the reason why when cooking a chicken at 140F in the oven in dry California the meat's temperature stalls at about 132F after several hours and doesn't budge again after even several more hours. (Heston Blumenthal's 'perfect chicken' is cooked to 140F in a 140F oven -- but I have never managed to get the interior temp of the meat up to 140F without cranking up the oven to something like 155F.

Nathan, what do you think?

Very interesting. So let me see if I get this right...

For a given piece of meat in a barbecue-style smoker the temperature of the meat will slowly rise until it reaches a kind of "temporary equilibrium temperature" where evaporating moisture maintains the surface of the meat at a more or less constant temperature via the "wet bulb effect" (the inside of the meat also plateaus since peak internal temperature is dependent on surface temperature). The temperature of this temporary plateau should be somewhat different depending upon the ambient temperature and humidity in the smoker, but I imagine these variables are similar enough in barbecue-style cooking to account for the broad similarity observed in this temperature plateau. So the meat sits at the plateau temperature until the surface of the meat dries out sufficiently so that evaporation is no longer able to cool the surface of the meat effectively in the ambient environment (this is now "dry bulb"), at which point the surface temperature and consequently the internal temperature can rise. The time that the meat spends at the "temporary equilibrium temperature" until the surface dries out is the observed "temperature stall."

That is not only incredibly interesting, but also a fun debunking of what everyone thought was a revelatory explanation of the science behind this observed effect! I wonder the extent to which the "temporary equilibrium temperature" could be altered by introducing moisture into the smoking chamber.

Oh, and don't worry. I am most certainly buying the book!

[PedroG]

Some time ago I did an experiment showing extreme evaporative cooling:

I soaked a wet pile of rags about 10cm thick with water of 50°C and placed it in my convection oven set to 50°C. One temperature sensor was placed in the center of the rags, the other was placed in the oven near the pile. Evaporative cooling made the core temperature of the package drop to 36°C, then I lost my patience and wrapped the pile in cling-film to stop evaporation (which corresponds to completely drying the surface of the meat by blow-torching), and temperature rise to the final core temperature took about 6 hours which is in the 4-8h range described by Heston Blumenthal in his non-SV two-bone-fore-rib (From The Sunday Times, November 12, 2006, "Steak with blue-cheese-infused butter and mushroom ketchup"). Evaporative cooling not only kept the core temperature lower, but also lowered the oven temperature.

[nathanm]

I wonder if this is the reason why when cooking a chicken at 140F in the oven in dry California the meat's temperature stalls at about 132F after several hours and doesn't budge again after even several more hours. (Heston Blumenthal's 'perfect chicken' is cooked to 140F in a 140F oven -- but I have never managed to get the interior temp of the meat up to 140F without cranking up the oven to something like 155F.

Nathan, what do you think?

That is exactly what is going on. To cook a chicken to a temp of 60C/140F, you need the wet bulb temperature in the oven to be that high. If there is 100% relative humidity then wet bulb temp = dry bulb temp. So you need to either introduce more humidity, or raise the dry bulb temperature.

[FoodMan]

So, in a 100% humid envrionment, we will not observe a stall but we will observe a much gradual temperature increase over a longer period of time depending on the size of the meat?

[nathanm]

We did the stall experiment. We took a brisket, and cut it in half - one piece was sous vide vacuum packed, the other not. Both pieces were put in a combi-oven in convection mode at 90C/194F. We had temperature probes in the air, in the core, and at the surface of each one.

The stall occurred for the un-bagged brisket. The surface temperature of the bagged and unbagged meat started to diverge at 72C/161F. Below that point the bagged one was slightly lower (due to insulation effect and slick surface of plastic), but they crossed over at 72C and from that point forward the bagged one was hotter (both surface and core).

The flat part of the stall was at 78C/172F. The flatest part of the stall curve lasted about 2.5 hours, then gradually creeped upwards. The temperature of the un-bagged brisket was still lower than the bagged brisket after 5.5 hours.

I think this conclusively shows that wet bulb temperature effects (i.e. oven humidity and drying) are the source of the "stall" in doing BBQ. Uncovered meat will stall due to evaporative cooling, which gradually goes away as the surface dries.

This also points something out that I am not sure everybody knows - you can cook sous vide in a convection oven. The reason that we all use water baths is that most convection ovens have poor temperature control compared to a water bath. The heat transfer rate to the food is also much worse, because water is about 1000 times denser than air. However you can do it. In this case we did it because I wanted the two brisket pieces to be under the same temperature and heat transfer rate - the only difference was surface evaporation.

[nathanm]

So, in a 100% humid envrionment, we will not observe a stall but we will observe a much gradual temperature increase over a longer period of time depending on the size of the meat?

Yes - a 100% humidity environment is what you have inside a sous vide bag, or in a covered pot, or in a combi-oven in steam mode or CVAP oven. In that case there is no evaporative cooling, and there is no stall.

[jackal10]

Does the crust formation, besides cooling, affect flavour and texture?

I can imagine it would cocentrate salts, and also change the condensation/adsorbtion of compounds from the smoke.

Conventional wisdom in cold smoking requires a dry surface pellicule.

In other words why don't BBQ cooks wrap their meat or pre-dry it, for example overnight in a fridge?

Edited August 19, 2010 by jackal10 (log)

[Chris Hennes]

In other words why don't BBQ cooks wrap their meat or pre-dry it, for example overnight in a fridge?

To avoid stall? Is there a reason it might be desirable to avoid it? Just faster cooking?

[DouglasBaldwin]

It's been quite awhile since Nathan or I got into the science of heat transfer in food on here. Since Nathan's recent posts on `stall' have a lot of science behind them, I'd like to unpack them a bit. I hope my expanded explanation from a slightly different perspective will aid your understanding.

Heat transfer in food is wonderfully complicated! If it were simple, there wouldn't be several academic journals devoted to its study. First I'll discuss how Nathan and I model heat transfer in sous vide cooking. Then I'll discuss the more complicated problem of modeling heat transfer in dry ovens and smokers. [i won't discuss the mathematics because it's beyond the scope of this thread. I'll restrict my discussion to meat.]

SOUS VIDE COOKING

We can very accurately model the transfer of heat in sous vide cooking if we know:

• the thermal properties of the food,
  
• the food's shape and initial temperature,
  
• the surface heat transfer coefficient of the water bath (or convection steam oven), and
  
• the water bath's temperature.
  

For example, Nathan emailed me some measurements he took while cooking a 80mm cylinder of beef and my mathematical model calculated the core temperature (without any fiddling with the parameters) to within +/-0.1F!

Thermal Properties of Meat

We model the temperature inside the meat using the heat equation. The only parameter in the heat equation is the thermal diffusivity of the meat.¹ Between refrigerator temperatures and boiling, we usually assume that thermal diffusivity is constant even though it change slightly with temperature.

The change in thermal diffusivity with temperature depends on the specific heat capacity of the meat. As Nathan mentioned, one way of measuring the change in specific heat capacity is with a differential scanning calorimeter (DSC).² The specific heat capacity tells how the temperature changes with the addition of energy. Kemp et al. (2009) showed that the specific heat capacity of muscle fibers and sarcoplasmic proteins is essentially flat from 30C to 85C, connective tissue has a hump peaking around 60C, and fat is essentially flat over 50C. Therefore, the `stall' in smoking at around 74C isn't caused by chemical changes in collagen or the melting of the fat.

There are huge change in thermal diffusivity of meat outside 0C to 100C. Melting/freezing takes a huge amount of energy: it takes about as much energy to heat beef from -20C to 0C as it does from 0C to 70C. It takes even more energy boil away the water in the meat, but we don't have to worry this when cooking sous vide so I'll discuss this later.

Shape and Initial Temperature

See my previous posts about shape and cooking times. In short, the shape of the food has a huge affect on the cooking time. Initial temperature, so long as the meat isn't frozen, doesn't really matter.

Surface Heat Transfer Coefficient and Water Bath Temperature

To be able to calculate the temperature inside the meat (using the heat equation),³ we need to know how the temperature changes on the surface of the meat. In sous vide cooking, we can very accurately compute this from the temperature of the water bath and something called the surface heat transfer coefficient.

The water bath's temperature and how to control it has been discussed extensively previously in this thread, so I won't discuss it further.

The surface heat transfer coefficient is an intrinsic property of the water bath or convection steam oven.⁴ It describes how efficiently the energy in the bath can be transferred to the meat. In water baths, it mainly depends on the natural or forced convection (or movement) of the water. [in convection steam ovens, it mainly depends on the condensation of water vapor on the food.] If you've been outside when it's cold, you've probably experienced this first-hand: when it's calm out, your body is able to heat a thin layer of air around you so it doesn't feel so cold out; when it's windy out, the wind is continually removing this warmed layer of air so it feels much colder out. It's the same idea when cooking food in a water bath: if there isn't any convection, then the cold food cools a thin layer of water around it and the surface temperature takes longer to reach the water bath's temperature; if the water is always moving, then this layer of cooler water next to the food is removed and the surface temperature reaches the water bath's temperature much faster.

The surface heat transfer coefficient, h, is measured in W/m²-K. Once h is above a certain threshold, there is practically no difference in how quickly the core temperature increases. That is, once h is high enough, the limiting factor on how quickly the core temperature changes is the thermal properties of the meat. I've started doing empirical tests to measure h for all the water bath's we typically use. I've found that h is about 95 W/m²-K for my SVS and about 155 W/m²-K for my 2 year old PolyScience immersion circulator: both these h are above this threshold. In dry ovens (both convection and standard) and consumer smokers, h is below this threshold and small difference will cause large changes in the cooking time --- Nicolaï and Baerdemaeker (1996) reported that the h of dry convection ovens is about 14--30 W/m²-K. So different ovens and smokers will have significantly different cooking times.

DRY OVENS AND SMOKERS

While the meat at the surface still retains most its water, we can use the same heat equation and take the thermal diffusivity to be constant. The main differences are that h is much lower and how quickly the surface temperature increases is proportional to the difference between the wet-bulb temperature in the oven and the current surface temperature of the meat. (In the water bath, the surface temperatures rate of change was proportional to the water bath's temperature, which is always equal to the wet-bulb temperature, and the surface's current temperature.) As Nathan's explained, the wet-bulb temperature is cooler than the dry-bulb temperature in an oven when the relative humidity is less than 100% because of evaporative cooling. Evaporative cooling is so important because it takes a huge amount of energy to change the phase of water to water vapor --- it takes five times as much energy to evaporate a mass of water as it does to heat it from 0C to 100C!

When the meat at the surface has desiccated (dried out), we can no longer assume that the thermal diffusivity is constant --- indeed, it changes dramatically. This makes it very difficult to calculate the core temperature of the meat and so is beyond the scope of this post. This is also when, as Nathan discussed, that the dry-bulb temperature in the oven or smoker becomes more important. Since this doesn't really apply to sous vide cooking, I'll leave it for a later discussion; if you're really interested in how we deal with this `moving boundary value problem' then either Nathan or I can direct you to the relevant literature or discuss it on a more relevant eGullet thread.

Footnotes

1. Thermal diffusivity is defined as the thermal conductivity divided by the specific heat capacity and the density.
   
2. The very small sample size that must be used in DSC, usually less than 0.1 grams, makes getting a representative sample difficult. So most studies look at the meat's components --- myofibrillar proteins, sacroplasmic proteins, connective tissue, and fat --- separately. Previously, adiabatic calorimeter were used because they accept much larger sample sizes.
   
3. The heat equation is very well behaved, if we know the temperature on the surface and the initial temperature, then there is always a unique solution. If the water bath is held at a fixed temperature, then the heat equation tells us that the interior of the meat will (eventually) go to that temperature.
   
4. How the food is arranged can detrimentally affect the surface heat transfer coefficient:. Closely packed food cannot be heated as efficiently as adequately spaced food, which is why it's important to keep the food pouches well separated.
   

References

• Robert Kemp, Nike North, and Shane Leath. Component heat capacities for lamb, beef and pork at elevated temperatures, Journal of Food Engineering (2009) 92, pp. 280--84.
  

[PedroG]

We did the stall experiment.......

Would you mind posting the graph? Thank you.

[nathanm]

We did the stall experiment.......

Would you mind posting the graph? Thank you.

We repeated it again last night, with the same result. I will post a graph when I get the time.