Posts
This is part 3 of my analysis of data from my first full series of creaming experiments with Weyauwega butter. Here is the data, again - column headers would be:
minutes - start temp F. - end temp F. - wt. ingred. + cup(g) - Wilton calib time
0 - NA – NA – 312 - NA
1 – 69.0 – 68.5 – 280 – 7.5
2 – 70.7 – 70.3 – 248 – 7.5
3 – 68.6 – 69.4 – 243 – 7.25
4 – 70.0 – 70.7 – 227 – 7.25
5 – 70.2 – 71.3 – 225 – 7
6 – 68.7 – 71.1 – 236 – 7
8 – 69.5 – 71.5 – 220 – 7
10 – 69.8 – 74.3 – 223 – 7.25
13 – 69.5 – 76.5 – 219 – 7.25
I am going to present the data in slightly different form, with column headers:
minutes - change in temp F. - wt. ingredients - %volume increase
0 - NA – 258 - 0
1 – -0.5 - 226 - 14.2%
2 – -0.5 - 194 - 33.0%
3 – 0.8 - 189 - 36.5%
4 – 0.7 - 173 - 49.1%
5 – 1.1 – 171 - 50.9%
6 – 2.4 - 182 - 41.8%
8 – 2.0 - 166 - 55.4%
10 – 4.5 - 169 - 52.7%
13 – 7.0 - 165 - 56.4%
First, let's look at the change in temperature during the creaming process. The temperature drop in the first two times surprised me. I thought maybe I'd recorded the wrong starting temperature when I saw the ending temperature of the one-minute time was lower than the start, but then the same thing happened during the two-minute time. The most logical explanation might be that either the bowl or the sugar was colder than the butter so when I started beating the butter it cooled the butter down, but in fact everything had been sitting out for at least several hours, and nothing should have been warmer or colder than anything else. I think it's pretty interesting, but I really cannot explain it except by wild guesswork. My wild guess is that the energy content of butter that has not been beaten is lower than the energy content of butter that has been beaten, due to some sort of molecular interaction. When the butter is beaten, energy is lost to the material in order to break up the molecular interactions, and that energy loss shows up as a small drop in temperature. I wouldn't even believe that tentatively, though - it's a total guess, and who knows what the real reason is? Anyway, I think the temperature drop is real and simply accept it as an observation. It's quite a small drop and has little effect on the end result. If the temperature actually increased with each time right off the bat, it would be bad news. Luckily, it does not.
There is a clear temperature rise as beating continues. The dough after ten and thirteen minutes creaming was really gooey, so much so that the only way I could form it into rolls at all was to place it on the wax paper and pull it into an elongated shape, and then hold the two opposite edges of the wax paper up so that the dough was hanging in the air like it was in a hammock; then drop it a bit, abruptly, so that the bottom would take on a round shape. Then I wrapped the sides of the wax paper over the top and rolled it a little, gently, inside the wax paper.
If I was not trying to keep a set procedure I would have simply refrigerated the dough after 10 and 13 minutes' creaming for a while to let the temperature drop a few degrees, but I didn't want to deviate from the procedure.
The experiment I posted on January 29, 2010 suggested that I should have been able to cream the butter long enough to increase the volume by about 64%. However, my earlier experiment took 21 minutes to whip that much air into the butter and sugar. It was simply impossible to cream the butter and sugar for 21 minutes non-stop - I would have had nothing but a warm greasy mess on my hands. If I wanted a volume increase of 64% I'd need to have stopped after ten minutes or so, let everything cool back down, and then finished up. But I got around 55% increase by 8-10 minutes, and I would think that's good enough.
Another option to reach maximum volume would be to use the small bowl for creaming, and then transfer the creamed butter and sugar to the large bowl for mixing with flour. Creaming in the small bowl went much more quickly (see also the post of January 29, 2010).
One of the salespeople at William-Sonoma told me that the sort of wire beaters that are used on most mixers these days are better at whipping air into whatever they are beating. I don't know why that is, but maybe they could get the volume increase up to 64% or even higher, perhaps without as much temperature increase. But again, I should think 55% is sufficient.
The volume increase at 6 minutes was way out of line with surrounding times. When I weighed the cup after that time, I thought "What?" and double-checked the weight. If it was too light I'd have suspected a big air bubble beneath the surface of the butter in the cup, but since it was "too heavy" that wasn't it. The outside of the cup was clean. That was simply what it weighed, and I have no idea why. I keep thinking maybe I made a mistake in measuring the ingredients or something, but when I reviewed what I had done in my mind, I didn't recall a possible error in measurement. Possibly it would be worth tossing that number and repeating the 6-minute experiment. But that's a slippery slope! For all I know, the 4, 5, and 8-minute weights should have been greater. I think I need to just take the value as is for now.
My intention is to get triplicate measurements of volume after creaming for each time, so I can calculate an average and standard deviation. I will do another series, but there is no point of baking all the same sad cookies again. I think I'll create a slightly different procedure that is different only after the creaming step, so I can use the data from creaming for calculating an average, and still move forward with my experiments. I need to decide what change to the procedure might be worth running the entire time series on.
Last, a bit about the attached photo. In addition to baking each cookie on the Wilton pan, I decided to also bake them on the Williams-Sonoma Goldtouch Commercial Quality Cookie Sheet for around 12.5 minutes, starting with the 2-minute creaming time. The photo is a slightly blurry shot of these cookies. You can see these are not butt-ugly - they actually look nice. These cookies were also pretty flat, but not quite as flat as the butt-ugly cookies. The 8-minute cookie got a bit browned on one edge, but otherwise they are all pretty light-colored. There is no pitting on any of them, and they are all nicely baked. The bottoms show that obvious gas bubbles formed, but they were not as large in size as when the cookies were baked on the Wilton pan.
It's apparent in the picture that the 2, 3, 4, and 5-minute cookies had successively larger diameters. The same pattern shows in the cookies baked on the Wilton pan, but it is accentuated on the Goldtouch pan. The shorter-time cookies spread more on the Wilton pan, which is why the pattern was less strong there. On the Goldtouch pan, the poor shape of the rolls at the very high creaming times shows more, because there is less spreading. These rolls are fairly misshapen because their high temperature didn't allow them to hold their shape after they were placed in the refrigerator.
This picture constrasts strongly with the picture of the cookies baked on the Wilton pan. At all creaming times, the Goldtouch pan made far better-looking cookies. I am still looking for the key that will let these cookies look good on a hotter, non-stick pan. I think it's probably possible to make them perfectly well on non-stick pans, but these two photos certainly show that baking sheets with traditional finish and longer baking times can offer advantages when baking cookies with high butter content.
minutes - start temp F. - end temp F. - wt. ingred. + cup(g) - Wilton calib time
0 - NA – NA – 312 - NA
1 – 69.0 – 68.5 – 280 – 7.5
2 – 70.7 – 70.3 – 248 – 7.5
3 – 68.6 – 69.4 – 243 – 7.25
4 – 70.0 – 70.7 – 227 – 7.25
5 – 70.2 – 71.3 – 225 – 7
6 – 68.7 – 71.1 – 236 – 7
8 – 69.5 – 71.5 – 220 – 7
10 – 69.8 – 74.3 – 223 – 7.25
13 – 69.5 – 76.5 – 219 – 7.25
I am going to present the data in slightly different form, with column headers:
minutes - change in temp F. - wt. ingredients - %volume increase
0 - NA – 258 - 0
1 – -0.5 - 226 - 14.2%
2 – -0.5 - 194 - 33.0%
3 – 0.8 - 189 - 36.5%
4 – 0.7 - 173 - 49.1%
5 – 1.1 – 171 - 50.9%
6 – 2.4 - 182 - 41.8%
8 – 2.0 - 166 - 55.4%
10 – 4.5 - 169 - 52.7%
13 – 7.0 - 165 - 56.4%
First, let's look at the change in temperature during the creaming process. The temperature drop in the first two times surprised me. I thought maybe I'd recorded the wrong starting temperature when I saw the ending temperature of the one-minute time was lower than the start, but then the same thing happened during the two-minute time. The most logical explanation might be that either the bowl or the sugar was colder than the butter so when I started beating the butter it cooled the butter down, but in fact everything had been sitting out for at least several hours, and nothing should have been warmer or colder than anything else. I think it's pretty interesting, but I really cannot explain it except by wild guesswork. My wild guess is that the energy content of butter that has not been beaten is lower than the energy content of butter that has been beaten, due to some sort of molecular interaction. When the butter is beaten, energy is lost to the material in order to break up the molecular interactions, and that energy loss shows up as a small drop in temperature. I wouldn't even believe that tentatively, though - it's a total guess, and who knows what the real reason is? Anyway, I think the temperature drop is real and simply accept it as an observation. It's quite a small drop and has little effect on the end result. If the temperature actually increased with each time right off the bat, it would be bad news. Luckily, it does not.
There is a clear temperature rise as beating continues. The dough after ten and thirteen minutes creaming was really gooey, so much so that the only way I could form it into rolls at all was to place it on the wax paper and pull it into an elongated shape, and then hold the two opposite edges of the wax paper up so that the dough was hanging in the air like it was in a hammock; then drop it a bit, abruptly, so that the bottom would take on a round shape. Then I wrapped the sides of the wax paper over the top and rolled it a little, gently, inside the wax paper.
If I was not trying to keep a set procedure I would have simply refrigerated the dough after 10 and 13 minutes' creaming for a while to let the temperature drop a few degrees, but I didn't want to deviate from the procedure.
The experiment I posted on January 29, 2010 suggested that I should have been able to cream the butter long enough to increase the volume by about 64%. However, my earlier experiment took 21 minutes to whip that much air into the butter and sugar. It was simply impossible to cream the butter and sugar for 21 minutes non-stop - I would have had nothing but a warm greasy mess on my hands. If I wanted a volume increase of 64% I'd need to have stopped after ten minutes or so, let everything cool back down, and then finished up. But I got around 55% increase by 8-10 minutes, and I would think that's good enough.
Another option to reach maximum volume would be to use the small bowl for creaming, and then transfer the creamed butter and sugar to the large bowl for mixing with flour. Creaming in the small bowl went much more quickly (see also the post of January 29, 2010).
One of the salespeople at William-Sonoma told me that the sort of wire beaters that are used on most mixers these days are better at whipping air into whatever they are beating. I don't know why that is, but maybe they could get the volume increase up to 64% or even higher, perhaps without as much temperature increase. But again, I should think 55% is sufficient.
The volume increase at 6 minutes was way out of line with surrounding times. When I weighed the cup after that time, I thought "What?" and double-checked the weight. If it was too light I'd have suspected a big air bubble beneath the surface of the butter in the cup, but since it was "too heavy" that wasn't it. The outside of the cup was clean. That was simply what it weighed, and I have no idea why. I keep thinking maybe I made a mistake in measuring the ingredients or something, but when I reviewed what I had done in my mind, I didn't recall a possible error in measurement. Possibly it would be worth tossing that number and repeating the 6-minute experiment. But that's a slippery slope! For all I know, the 4, 5, and 8-minute weights should have been greater. I think I need to just take the value as is for now.
My intention is to get triplicate measurements of volume after creaming for each time, so I can calculate an average and standard deviation. I will do another series, but there is no point of baking all the same sad cookies again. I think I'll create a slightly different procedure that is different only after the creaming step, so I can use the data from creaming for calculating an average, and still move forward with my experiments. I need to decide what change to the procedure might be worth running the entire time series on.
Last, a bit about the attached photo. In addition to baking each cookie on the Wilton pan, I decided to also bake them on the Williams-Sonoma Goldtouch Commercial Quality Cookie Sheet for around 12.5 minutes, starting with the 2-minute creaming time. The photo is a slightly blurry shot of these cookies. You can see these are not butt-ugly - they actually look nice. These cookies were also pretty flat, but not quite as flat as the butt-ugly cookies. The 8-minute cookie got a bit browned on one edge, but otherwise they are all pretty light-colored. There is no pitting on any of them, and they are all nicely baked. The bottoms show that obvious gas bubbles formed, but they were not as large in size as when the cookies were baked on the Wilton pan.
It's apparent in the picture that the 2, 3, 4, and 5-minute cookies had successively larger diameters. The same pattern shows in the cookies baked on the Wilton pan, but it is accentuated on the Goldtouch pan. The shorter-time cookies spread more on the Wilton pan, which is why the pattern was less strong there. On the Goldtouch pan, the poor shape of the rolls at the very high creaming times shows more, because there is less spreading. These rolls are fairly misshapen because their high temperature didn't allow them to hold their shape after they were placed in the refrigerator.
This picture constrasts strongly with the picture of the cookies baked on the Wilton pan. At all creaming times, the Goldtouch pan made far better-looking cookies. I am still looking for the key that will let these cookies look good on a hotter, non-stick pan. I think it's probably possible to make them perfectly well on non-stick pans, but these two photos certainly show that baking sheets with traditional finish and longer baking times can offer advantages when baking cookies with high butter content.
No comments:
Post a Comment