another_jim wrote:Thanks, that makes a lot more sense.

But I still think that the jiggles in the RoR plots are mostly noise, i.e. not electric noise, but local and transient changes in air temperatures around the BT sensor, rather than anything in the bean mass as a whole. The only way to actually test this argument would be to install a second bean temp sensor, and compare the two RoR plots. If the jiggles in the two plots correlate, I'm wrong, if they don't, I'm right. Any volunteers?

Dustin Demers did this years ago. I remember him talking about it a lot, but I don't recall him making any hard, one-sided opinions. He was just obsessed with collecting as much accurate data as possible.

My intuition leads me to believe you're correct however.

To elaborate more, he'd remove most of the screws that held the front plate to the rest of the roaster and replace them with thermocouples. He and I both fiddled with our Quest M3 roasters. I stopped at 4 I think. But I clearly recall him using two descrete BT probes on his early Santooker roaster. Some of his old posts on here, if they show an Artisan graph, may show these.

Hmmm, I could easily do that on my Cormorant. There is a hole for an analog gauge that I never use and I have extra TCs and empty ports on my phidget. They are spaced about 1/2" apart so they would read difference areas of the bean mass. That would be interesting. Then I could see two different RoRs to see if they correlate. However, I'm more concerned with general trends rather than small changes, as stated before.

An array of IR temp sensors could be interesting too. Almost like thermal imaging like a FLIR provides. Surely someone has done that to the bean mass?

A couple notes:

I'm one of those people who has put probes everywhere. My first career was in nuclear physics assisting designers of reactor control and shutdown systems, and I was pleasantly surprised that my 1970's knowledge of sensor placement and treatment of sensor signals in a nuclear reactor could be applied to conventional gas drum coffee roasters.

OTOH, I have no idea how all these different probe measurements and associated statistics can be used to affect flavour. I've removed most of the probes, and mostly use the more unusual sensors like faceplate temperature and intake temperature to better set consistent starting conditions for roasts ("between batch protocol" in Rao-speak).

One comment on "declining RoR." Some insist that the RoR decline linearly - a straight line. Some say linear from Dry End onwards. At least in my roaster, I do not concern myself with linearity. If I run a 4mm diameter probe bolted to the faceplate, then I can get a linear straight-line RoR curve. Decouple the same probe from the faceplate with insulation, or run a 2.5mm insulated probe, and those RoR curves will not be linear for the exact same roast, although the curves must always be declining. I have asked Rao about this in-person, and he also posted about it a few years ago. To paraphrase, if you use his probe recommendations in the appropriate roaster, than a linearly-decreasing RoR may be the desired ancillary result. If not, then the curve may not be linear.

All I know for sure is my roaster (Bella Taiwan) has a reputation for being very easy to roast on, and it seems about as easy as driving a Toyota Camry. I've never had a 'bad' roast, even when I've tried things like turning off the gas at first crack and just coasting to drop. This is very different from my Hottop, which seemed to give baked and/or under-developed roasts unless I was 'perfect' with my power and airflow changes.

Milligan wrote:
I feel like it all links together and I don't believe Rao ever said RoR is an end-all-be-all but an indicator of a good roast.

I've long felt like Rao threads the needle between claiming that the perfect RoR is the end-all-be-all and saying that RoR curve isn't everything. He does, after all, sell a service and cultivates a reputation. His "safe area" to fall back on is the idea that an ideal declining RoR is easiest for repeatability.

For those who have not watched Rao's video on RoR as well as bean prob locations.

Jim,

could you help a non-native speaker to understand the paragraph below?

another_jim wrote:In my roasting model, I allow the thermal mass of the beans to change over time, so that the same amount of heat can change their temperature more as the roast progresses. When I fit the data to determine this, there is a fairly drastic step change at the outset of the first crack, when the beans start steaming, otherwise the thermal mass stays constant. The added activity in the RoR traces just ahead of the first crack may indicate such a change. On the other hand, the RoR amplitude also picks up at the yellowing stage, where my roasting models do not indicate sudden thermal mass changes

Typically, the specific heat -- the number of degrees 1 gram rises given the input of some 1 calorie of heat -- is a constant. But as coffee beans lose water, the number of calories required to raise their temperature also drops. If you are modelling a roast, the coffee beans' specific heat is not just one parameter to be estimated, but something changing over time. In my model, the most parsimonious good fit was using two specific heat values, a higher one for before the first crack, a lower one during and after the first crack.

This change explains the tendency of the bean temperatures to rise quickly in the first crack, and the need to drop the heat coming into the first crack. I was commenting that many of the ROR graphs show a similar tendency at around 150C, at the yellowing point, which was not reflected in the data sets I used for my model.

luca wrote:Well

Yeah this is something that I find mindblowing about people's speculation about coffee roasting. Clearly there's water being driven to steam, and I see all of these generalisations about how the BT will asymptote out towards the ET naturally. But all of this assumes the above relationship is all that is relevant. But it's not the case. There's a massive amount of energy required for the phase change from liquid water to steam:

https://en.wikipedia.org/wiki/Enthalpy_of_vaporization

I haven't really done any numbers or anything, but I'd have thought that this would account for something in the system eg. some non-linear behaviours in the BT approaching ET.

Maybe the amount of water is small enough, and the phase change happens staggered enough through the roasting process, that it sort of doesn't result in too much of a deviation from the generalised specific heat capacity mode.

I touched on that in post #14 of this thread. Coffee is certainly not a simple system. The material changes physical properties throughout the roast. One that isn't discussed as much is the discernible difference in temperature rise after first crack end. Many roasters completely cut gas (or lower it to 5% or so) and still coast for 20-30F to get into 2nd crack quite easily.

another_jim wrote:Typically, the specific heat -- the number of degrees 1 gram rises given the input of some 1 calorie of heat -- is a constant. But as coffee beans lose water, the number of calories required to raise their temperature also drops. If you are modelling a roast, the coffee beans' specific heat is not just one parameter to be estimated, but something changing over time. In my model, the most parsimonious good fit was using two specific heat values, a higher one for before the first crack, a lower one during and after the first crack.

This change explains the tendency of the bean temperatures to rise quickly in the first crack, and the need to drop the heat coming into the first crack. I was commenting that many of the ROR graphs show a similar tendency at around 150C, at the yellowing point, which was not reflected in the data sets I used for my model.

it's basically this part above and the fact that need for power input decreases that brought me to my current 'system' of keeping air flow constant, gas power input steady for x seconds (3'20 or thereabout) from charge (no soak) to then decrease by Y Kpa per Z seconds (0.5KPa per 30 sec). The specific heat of my roaster is constant, that of the beans is not. It seems to work in a perf drum Huky doing 400g batches pretty nicely.