Saturday, January 27, 2007

Steering with the throttle

In the middle of an interesting phone conversation with Philip Search, I was trying to find an analogy to help explain a rather obscure aspect of espresso making. I hope it didn't throw him off too much when I came up with the following:

If you're ever done automobile racing, you've experienced the phenomenon of steering with the throttle. When you're in a curve, and your tires are working near their limit of adhesion, you can actually keep the steering wheel fixed and steer the car with the accelerator pedal. It is a revelation to experience this for the first time, because nothing in everyday driving tips you off that it might be possible.

By the way, please don't attempt to learn how to do this without an instructor on a track. Otherwise, you may experience more than you bargained for and, as they say in racing, fail to keep the shiny side up!

OK, so what does this have to do with coffee? Some time ago, I was experimenting with espressos, trying to figure out how preinfusion (or the lack of it) affected the extraction. And I stumbled upon something unexpected.

I had rigged up a series of valves that allowed me to flip back and forth between two extraction scenarios. In one, the initial flow of water to the group was restricted so that the buildup to full 9 bar pressure took about 7-8 seconds. In the other, the initial flow was unrestricted and it took only a second or two to get 9 bars. The difference was roughly like having installed an 0.6mm gicleur (slow buildup) and then switching to a 1.0mm or larger gicleur (fast buildup).

The surprise was that the fast pressure buildup had the same effect on shot timing as if I had used a much finer grind: the shot flowed so slowly from the portafilter that I had to grind more coarsely for subsequent shots. This was the opposite of what I had expected (which was that the faster pressure buildup would result in a shot that took less time).

At the time, Dr. John offered an explanation for why this occurred, and he's probably correct. But the bottom line is, not only does preinfusion lessen the chance that you'll have channeling in the portafilter, it also changes the grind required for your desired shot timing. Exactly how this affects the shot's flavor and texture is a vast subject....

I always thought this was pretty interesting; I hope you do, too. So, kids, be open to new phenomena, experiment with preinfusion in your espresso making, and always keep the shiny side up! :-)

19 comments:

  1. Dear Gicluer specilist... I recall a post somewhere (coffeed.com me thinks) a post highlighting that most (/all?) commercial Italian Espresso Machines had a pre-infusuon/ gicleur spec'd which is best suited to blends using a portion of Robusta; that the pre-infusuion was longer and not best suited to wholly Arabica blends? Has this stood up (i.e. proven) and is it something worthwhile addressing on a Cimbali machine? Is gicleur surgery warranted if I am only using 100% Arabica blends?!

    ReplyDelete
  2. VR6??? Hmmm, me thinks the 86 8v GTI and I need to take a tour up to New York...

    Interesting post Andy. I've been meaning to experiment with a similar phenomenon that occured quite accidentally this summer.

    In midst of shot pulling, the pump motor shuts off about say, 15 seconds into the pull. So the shot starts off by ramping to 8.5 within the first second (standard AV), then drops down to line pressure (6) when the motor shuts off, leaving the last ten seconds to pull under line pressure.

    My initial tastings were decent but the machine was pulled from service shortly after the problem started and it's now sitting in my basement where I do not have 220v power set up.

    Any theories on how this may affect the shots?

    ReplyDelete
  3. I quite fancy experimenting with this, but what is the best way to go about it?

    ReplyDelete
  4. I agree steering with the throttle is very reel, but with a front drive Scirocco? Does it display this kind of oversteer capability?
    I've noticed this in preinfusion for a while, but if I read Dr. John's theory correctly, I don't agree with it. I think it's simply instant compression of the puck which brings all the grinds closer together slowing everything down. I really don't think the fines have time to migrate to the bottom. Then again, it's just another theory of mine.
    John

    ReplyDelete
  5. Hi Anonymous, I don't know anything about preinfusion designed specifically for Robusta blends. Many people find they prefer the results of gicleured machines with their Arabica blends.

    Try removing the portafilter from one of the groups on your Cimbali, then turn on the pump for exactly 10 seconds. Measure how much water flows from the group. If it's in the range of 60-90ml, you don't need "gicleur surgery." If it's above 90ml, you might consider the operation, which is usually done under a local anesthetic and may even be covered by your health insurance.

    ReplyDelete
  6. Hi Jay:

    Sadly, I traded in the Corrado VR6. The picture was taken by the dealer as I said goodbye to the car, which is why I have a wistful look on my face.

    If your shot started out at 8.5 bar and then drifted down to 6 bar when the pump cut out, you had a pressure profile pretty similar to a lever shot. No reason why it couldn't have been delicious.

    ReplyDelete
  7. James, I'm not sure how to do this on your machine, but here's a pic of the valve setup I used to switch from low flow to high flow:
    http://i17.photobucket.com/albums/b62/shekk/debit5.jpg

    ReplyDelete
  8. Hi John:
    Actually it was a Corrado, not a Scirroco, but FWD nonetheless. Although you can't kick out the rear end by stabbing the throttle on a FWD car, the throttle steering near the limit is still very effective. It's probably more a function of front tire traction than weight transfer in a FWD.

    ReplyDelete
  9. Your analogy didn't throw me Andy, it was great. I hope I didn't like too much of an idiot as I was abit tired. Really enoyed conversing though.

    ReplyDelete
  10. Andy~ Very interesting analogy. This means I am now going to have to spend even more time in front of my Synesso pulling shots with many different preinfusion times...

    ReplyDelete
  11. Hey Philip: I too thoroughly enjoyed the conversation, it could have gone on for hours.
    Hi Ryan: Real interested to hear about your preinfusion experiments.

    ReplyDelete
  12. I think I agree with "Anonymous" aka John.

    Pre-infusion is effective because it evenly saturates the puck rather than hitting it like a ton of bricks and exploiting every weak spot under high pressure.

    Lower pressure allows the water to disperse more evenly in the puck. This water can ALSO help the coffee particles to shift slightly (like quicksand, but FAR less extreme) to actually aid in distribution of grounds, and thus flow resistance, before the pressure is applied.

    -Jason "water is lazy" H.

    ReplyDelete
  13. I likewise agree with jason and "anonymous". Pre-infusion simply prepares the puck for extraction, allows the water to flow through instead of further compressing the puck. Think about sponges, if you hold a sponge and spray it with the hose, the water will always flow through the sponge and out the bottom more easily if the sponge is already wet.

    Also, the idea of having large particles to control flow, and small particles for flavour just seems bizarre. The reason you get better extractions when you have some small particles and some larger, is that they mesh together better in the puck. If you have a whole bunch of pebbles (of a roughly equal size) in a jar, you tend to have a lot of air in between, whereas if you have some smaller stones mixed in, the empty space is much less.

    ReplyDelete
  14. John said:
    "if I read Dr. John's theory correctly, I don't agree with it. I think it's simply instant compression of the puck which brings all the grinds closer together slowing everything down. I really don't think the fines have time to migrate to the bottom."

    Not sure I understand what you're saying, but the fines certainly do have time to migrate to the bottom -- Illy presents good evidence of this in his chapter on "percolation."

    Jason said:
    "Pre-infusion is effective because it evenly saturates the puck rather than hitting it like a ton of bricks and exploiting every weak spot under high pressure."

    Yeah, that's the conventional concept, but it's not the whole story.

    "Lower pressure allows the water to disperse more evenly in the puck. This water can ALSO help the coffee particles to shift slightly (like quicksand, but FAR less extreme) to actually aid in distribution of grounds, and thus flow resistance, before the pressure is applied."

    But what you say is the OPPOSITE of what I found. Effective preinfusion reduces resistance to flow, allowing a finer ground to be used.

    Phil said:
    "Also, the idea of having large particles to control flow, and small particles for flavour just seems bizarre."

    The idea is more complex than you state. The fine particles perform two functions: they provide high surface area, promoting extraction, and they plug holes, slowing down the flow rate (which also promotes extraction). The large particles provide a porous structure that doesn't get the flow rate TOO slow.

    Approximately 9 bars of pressure is necessary to properly extract lipids and other insolubles. At that high pressure the proper fine/coarse balance is necessary to keep the flow rate low enough so that there is time to properly extract the coffee.

    ReplyDelete
  15. I think I agree with you really. What I said was merely simplified for conciseness, possibly didn't convey my real meaning.
    My point was the the coarser particles aren't there solely to contol flow rate, as you said, "the proper fine/coarse balance is necessary to keep the flow rate low enough so that there is time to properly extract the coffee."

    Also consider that if the finer particles are properly extracted, the coarser ones are mostly still under-extracted, and the correct balance of different particle sizes (and therefore different levels of extracted-ness) must play a part in having the correct balance of flavour in the resulting espresso.

    I really do have a problem with the idea of the fines migrating, but as I don't recall Illy's "percolation" chapter, and don't have a copy on hand, I'll do my homework first like a wise little monkey.

    ReplyDelete
  16. Andy, I think the fines do migrate during extraction(the entire extraction time), just not as I understand Dr. John to say they go to the bottom of the basket and cause an obstruction when pressure is applied suddenly. I just can't believe they can move that fast.
    I stick with my theory that the puck becomes a sort of diaphram valve that is instantly compressed, slowing flow. Actually I'm not sure that it matters why, and if anything can ever be generalized from this. I enjoy some coffees with a faster ramp profile, and some a slower one, to get a group of people to agree what taste best--never happen, that's what espresso has taught me, everyone has their own taste.
    I have done some crazy experiments, and always ultimately base what I find on taste, my taste, which is my favorite =) even though it may or may not be what everyone else likes best.
    John

    ReplyDelete
  17. John and Phil, I appreciate the comments.

    Phil, Illy cites electron microscopy that presumably shows the fines collecting near the bottom. Also, he had a trick portafilter/basket which could be flipped upside down and reinserted. He found that the flow was high for the first few seconds after turning on the pump, then it slowed down. When the portafilter was flipped, the flow again went high for a few seconds before slowing down. He attributed this to the supposition that the fines "counter-migrated" to the new, opposite bottom of the portafilter, where they resumed restricting the flow.

    ReplyDelete
  18. http://13181810.blog.hexun.com
    http://13181921.blog.hexun.com

    ReplyDelete