What determines the assist affinity on the HCH-II?

Discussion in 'Articles' started by msantos, Jul 14, 2008.

  1. msantos

    msantos Eco Accelerometrist

    What factors determine the amount of assist (if any) the IMA_system uses when accelerating?

    [xfloat=left]http://www.cleanmpg.com/photos/data/501/HCH-II_Review_-_Headline.jpg[/xfloat]Manuel Santos – CleanMPG – July 14, 2008

    2008 Honda Civic Hybrid (HCH-II).

    Most of the assertions I will include in this article will be mainly related to the IMA Gen4 (HCH-II) as its inputs and relative complexity has a few differences with the previous generations.

    As many have noticed the electric assist routines on the HCH-II can change at times to the point of almost seeming unpredictable. For many HCH-II owners this can be as puzzling as it can be frustrating leading many to believe that some sort of manual control of assist would be best and more helpful at least from a fuel economy perspective.

    But when we look at this assist issue more carefully, particularly in the context of some technical information, many patterns often evidenced by thresholds begin to emerge. I will try to identify those patterns and thresholds here in an attempt to explain the why's and how's of the assist routines.

    Ambient temperature

    Ambient temperature actually plays a very significant role and at times, more so than the other operational temperatures (ICE, Transmission, IPU, NiMH modules, fuel management and emissions control system). This is because it sets the stage for an off-set temperature window that affects the governance of the other system's own operational limits. Of course there are some systems that have very specific and rather inflexible operational constraints but in general their operational range is also wide enough for them not to be a primary constraint.

    So to keep it simple, ambient temperature is a determinant of the amount of assist. For example very low temperatures tend to call for an increase of regen affinity. The same is often observed for very high ambient temperatures. This behavior is by design and Honda can alter it with software updates.

    Engine Temperature

    Engine temperature can also have an impact on the amount of assist used lower speeds depending. But the main reason why the assist affinity is often increased is not so much to cool the battery modules, improve torque or improve fuel economy but rather to smooth the power delivery curve and help eliminate ICE roughness at certain modes.

    Honda has found that the application of moderate levels of assist during the engine warm-up phase can have a very positive impact on the overall levels of vibration as well as engine reliability. The lower amounts of stress due to active vibration management allowed for weight reductions and other optimizations factored into the ICE design which also lowered the manufacturing costs (every little bit, however small counts).

    State of Charge

    Yes. The SoC plays a very important role in increasing the assist affinity. To keep it simple, lets us just say that the NiMH chemistry has a very characteristic charge curve and that a charge level at either end of the spectrum is neither optimal nor desirable. In other words, too high or too low a charge is not something the battery management routines will allow for very long and either an aggressive assist or forced regen is often used to bring back the SoC to the optimal operational range. This optimal range generally persists in the 60%-80% range (give or take a few percentage points).

    So, if we have 7 bars SoC or better then we'll have a very hard time preventing the overly aggressive assist. Likewise, if the SoC is at 4 or lower... or a forced regen pattern is underway, then preventing regen is harder to prevent. Depending on the SoC levels (and low module temperature) this regen pattern may only be somewhat avoided if the SoC reaches five bars or you switch the engine off and on again.

    This behavior is by design and Honda can alter it with software updates.

    NiMH Battery Module Temperature

    This is one of those inputs that is often overlooked. To keep it simple: When the module temps are too high the assist is VERY hard to avoid. A module discharge has the effect of cooling the pack - Go figure ;) Of course, there is also the SoC level acting as a limit before the IPU finally enters the dreaded managed mode.

    The reverse is also true. The BCM will prefer and call for extended periods of regen to warm up the pack in very cold days.

    This behavior is by design and Honda can alter it with software updates.

    Vehicle velocity

    The assist affinity will decrease as the velocity increases. There are several magical speed boundaries but the most important for most of us is 37km/h (23MPH). Beyond this threshold, the assist affinity is almost halved. This means that speeds up to almost 40km/h (25mph) require nerves of steel or some kind of a trick to avoid hitting EV assist. Beyond this speed marker it becomes easier to avoid it.

    I should also mention that "road grade" plays a very important role along with the speed in either enhancing or diminishing the assist affinity.

    This behavior is by design and Honda can alter it with software updates.

    Managed Mode or System Faults

    Managed mode is a containment (or protection sequence) that the system enters when the safe operational boundaries have been breached. This can happen in extremely cold days as well as extremely hot days. During this sequence, assist and regen can be significantly altered or even completely canceled. I was told that this managed mode can also be evoked by a system fault and that the engineering group was considering a possible separation of the two scenarios... I guess it really depends on how the platform ages and what challenges the field maintenance feedback will report on.


    At some point I'll be posting a couple of pictures of an assist inhibit module that I had installed on my HCH-II back in the fall of 2006. This module was quite unique in that it was one of just two of the same kind and also because it attempted to solve the assist problem.

    Later, I looked for other simpler and more affordable ways to influence the assist affinity, especially inhibiting it, even after it felt at times that it was more like trying to shoot a moving target especially after a new software update.

    As I found out a while later, I simply did not have enough interest from the HCH-II community in such a mod and it really made no sense for me to waste my time pursuing a simpler or more affordable solution.
    Also, as yet another blow, I also found that what I did was not only the "wrong approach" but also very expensive in more ways than one. But that is entirely another story. :(

    At the moment, my hopes for the strategic inhibiting of assist rests in the hands of the engineering teams working on the upcoming platforms. I just hope that I was convincing enough to get them to recognize the needs and wishes of the community.

    Until then we'll have to utilize whatever tricks and techniques we can devise and put them to the best use possible without undermining the systems or our warranties :)

    Anyhow, feel free to post whatever feedback, or assist inhibiting tips of your own to this thread.

    Cheers

    MSantos
     
    Last edited by a moderator: Oct 26, 2008
  2. pcope

    pcope Well-Known Member

    Re: What determines the assist affinity on the HCH-II ?

    hi MSantos,

    great info thread--thanks! Lots to digest. I've experienced the magic speed boundaries that you mention. Seems like there are also 'sweet spots' in the combined effects of ambient temp, SOC and vehicle velocity that can produce some great MPG, though I can see how recreating them on purpose might be challenging. In my case, it was hot but not boiling ambient temp, medium SOC, and fairly low/intermittent velocities-- so the assist and braking regen modes responded more sensitively than usual but surprisingly worked in my favor-- I ended up getting amazing (for me) milage. Your comments about the temp-dependent behavior of the NiMH battery and effects of ambient temp on the regen affinity help explain this (I think!).

    A quick question-- I've been reading about the software updates for the '07 HCH II, but when I asked the dealer about updates for the '08, they 'played dumb' (may have not been an act!). Do you by chance know if software updates for the '08 are already being released?

    thanks--

    Philip
     
  3. GreenVTEC

    GreenVTEC Well-Known Member

    Re: What determines the assist affinity on the HCH-II ?

    Wow some info :eek:

    I'm glad to see someone offering an explanation for some of the weirder things. I've noticed the car always tries to use assist going from a stop but somewhere in the 20's it stopped trying as hard. Nice to know 23 is a magic number I'll have to remember from now on!
     
  4. bill717

    bill717 Active Member

    Re: What determines the assist affinity on the HCH-II ?

    Thanks for the great explanation, from what you know about the IMA I would think you designed the thing :D!

    I had no idea there were so many factors that influence the amount of assist. I look forward to you posting some more info about your assist inhibit module, I for one would be very interested in such a mod.
     
  5. xcel

    xcel PZEV, there's nothing like it :) Staff Member

    Re: What determines the assist affinity on the HCH-II ?

    Hi Manuel:

    ___Is the BCM data available in the OBD-II stream to actually see where the SoC is vs. the non-discrete 8-pips on the SoC gauge? Second item, if you know where the SoC is and can maintain a 70% actual after removing the assist/regen with whatever trick is needed (consider the 00 – 04 Insight’s clutch mod switch) why not just let your SoC display let you control the protection vs. the multi-level protection circuit you designed for your HCH-II earlier? There has to be a way to disable the Assist/Regen action by interrupting a single data stream somewhere, right?

    ___I look at Honda’s NiMH protection as a safety for the average user that is going to be jumping on the pack during every accel and throwing a ton of regen at every slow down or stop. Since a hypermiler will not be driving in that fashion, maintaining the SoC within a tight band and throwing a switch to disable whatever bus is needed to remove regen/assist seems like the ideal to me?

    ___Good Luck

    ___Wayne
     
  6. msantos

    msantos Eco Accelerometrist

    Re: What determines the assist affinity on the HCH-II ?

    Hi Wayne;

    Yes, the SoC is available in the OBD-II stream since Honda's own HDS reads it without issues. The problem is that Honda has yet to release the details on how to retrieve many of the privileged values properly and they guard the information as if it was some sort of "State secret". I've asked for it and I did not get it. Anyway, we already spent a fair amount of time mapping as much of the PID landscape as possible and while there is a lot of encouraging data to read that is of potential value, there are also a lot of starting points and even more time to be spent deciphering it.

    From a hardware perspective, the "problem" for us is that the Gen 4 IMA system topology is now a network of specialized but yet cooperative/symbiotic systems. Gone are the days when you could simply interrupt/control a module with a simple sensor/line state/voltage hack and the other modules would happily carry on without questions. Nowadays, the modules not only check for the integrity of the communication patterns among them in the CAN buses, but they often correlate value ranges as part of known/expected states. Clearly, a hardware (brute force) solution is not the most attractive option - at least knowing what I know now. Judging from what the HDS does it is possible to even "write" to the modules, But again this is yet another one of those privileged functions that only the HDS software knows about. :(

    On a side note:
    Last fall, Ron (at LinearLogic) indicated that they were going to attempt the retrieval of Honda IMA specific PID's/data. At the time I also indicated/offered to send him whatever information I had that may help them expedite the research. Anyhow, since then I attempted to contact him a couple of times and I did not get much in terms of a reply, unless I am wrong (and I hate it when I am right) they may have encountered issues in this department as well. :(

    I agree on the rest.

    We definitely need more people working on this. ;)


    Cheers;

    MSantos
     
  7. Mendel Leisk

    Mendel Leisk 2010 Prius (CAN Touring) Staff Member

    Re: What determines the assist affinity on the HCH-II ?

    MSantos, if your don't mind, you could explain in even simpler language:

    "increase of regen affinity"

    in the context of:

    "So to keep it simple, ambient temperature is a determinant of the amount of assist. For example very low temperatures tend to call for an increase of regen affinity."

    In the above, are you saying the electric motor helps more, or less, in when it's very cold?

    BTW, excellent info, thanks! I'm going to go through it carefully.
     
    Last edited: Oct 26, 2008
  8. msantos

    msantos Eco Accelerometrist

    Re: What determines the assist affinity on the HCH-II ?

    Hi Mendel;

    Nice to see you here at CleanMPG... and Welcome !!!! :D :D

    Sure. Just like the "assist affinity" (which is an increased tendency to provide assist) the regenerative affinity (tendency to evoke regen) is governed by its own set of inputs and thresholds.

    For instance, when the temperatures are low the regen affinity increases. This means that the release of heat at the battery module level from increased regeneration has a "desirable" positive effect on the thermal management and health of the IPU and its battery pack. This may seem a little odd to some of us but it is a well known phenomenon that is factored in to the design and general governance of the NiMH pack.

    Now, there's a limit to how far down the temp scale this higher regen affinity will go. For instance, at temps below -18C the regen affinity is VERY VERY low - so low in fact that the only way to see a regen in some scenarios and when the engine is cold is to get the transmission in S while slowing down. Then again, at these low ambient temps the assist affinity is also low at least until the engine reaches a temp of at least 60C. This happens because at those low temps the tolerances are too low and may impact the "durability and longevity aspect" of the system as a whole.

    The opposite is also true. For instance, when it is very warm (hot actually) the regen affinity is a little lower. This is because lesser levels of regen help keep the pack temps lower. This lower regen affinity can be so low that it is very common for some of us to see an increase in SoC crashes which are not only due to several other (and more significant) factors but also because the lesser "tendency" to regen worsens (does not help) the SoC losses caused by the use of AC and other high drains.

    That is basically it in a - simplied - nutshell. Let me know if I was not clear enough. :)

    Cheers;

    MSantos
     
    Last edited: Oct 26, 2008
  9. Jess

    Jess Well-Known Member

    Re: What determines the assist affinity on the HCH-II ?

    "_I look at Honda’s NiMH protection as a safety for the average user that is going to be jumping on the pack during every accel and throwing a ton of regen at every slow down or stop. Since a hypermiler will not be driving in that fashion, maintaining the SoC within a tight band and throwing a switch to disable whatever bus is needed to remove regen/assist seems like the ideal to me?_"

    I agree 100%. I used to EV Glide every chance I got to achieve the highest mpg during my daily commute, but I found that the HCH II does not allow it without a battery drain and recharge. This takes a hit to the mileage erasing all the prior effort. Now, I'm realizing, that I have to take those EV Glide opportunities and drop a bar or two of regen to throw what I can back into the pack. This has worked out well because I rarely drain the battery and I can still average 65 mpg.

    It would be nice, though, to see exactly what the SoC is so one would know when they can take advantage of the EV Glide and when they should throw some regen back. Those battery pips just aren't accurate enough.

    Wonderful read, MSantos. Always an eye opener, when I read your insights into this car.

    Jess
     
  10. Harold

    Harold Well-Known Member

    Re: What determines the assist affinity on the HCH-II ?

    Do you get the SOC read-out with a CG? H
     
  11. Harold

    Harold Well-Known Member

    Re: What determines the assist affinity on the HCH-II ?

    Parden me ,I meant SG
     
  12. Harold

    Harold Well-Known Member

    Re: What determines the assist affinity on the HCH-II ?

    I mean SG not CG.H
     

Share This Page