It's been brought to may attention that you can't view
attachments on VAF if not logged in.
So I'm reposting here because I want to make the battery open
failure mode real to everybody. (I, for one, had not considered
that failure mode). Hope that those of you that have already
viewed this elsewhere will forgive me.
I did an autopsy (cut the lid off) on the
PowerSport YTX20L-BS 12V 20Ah AGM battery that failed open,
causing total loss of electrical power and engine stoppage
(electronic ignition and fuel injection) leading to forced
landing and damage to my RV-4 (separate threads elsewhere to
come).
Battery was mounted forward of firewall, so failure may be
heat and/or vibration related. Battery is designed for sports
vehicles, so should be vibration resistant. Could also be a
manufacturing defect. Battery was installed 29 Jul 2023. In
service for 45 hours with an estimated 50 cycles (starts).
Note that the battery did provide power to start the engine
with no problems 10 mins before the failure. Looking at engine
monitor log, about halfway into the flight spikes in voltage
started to show up and got bigger end bigger. So a somewhat
gradual failure.
Each of the six cells are interconnected via "bridges". I
assume zinc electrodes to lead electrodes. As you can see in
the second picture there is a crack in the lead bridge to the
second cell.
When I short the crack with a metal pick or connect the two
electrodes with wire clips the battery shows full voltage and
can easily deliver a 12 amp current. When not shorting the
crack, voltage drops to zero with even the lightest load (12V
0.16A fan).
I'll post a separate thread speculating on why the alternator
did not continue supplying the needed power.
In case you think this is a one-in-a-million case, an autopsy
was done in 2016 on an Odyssey PC680 battery with similar
failure mode -- again internal cell-to-cell interconnect
failure (third picture). Fortunately that guy had magnetos on
his Rotax.
My design philosophy was that the battery was the backup for
the alternator and the alternator the backup for the battery.
In my case that did not work out.
For any engine completely dependent (EFI and/or electronic
ignition only) my recommendation is to have a backup battery
behind the firewall. Connect the engine bus to either the main
battery or the backup battery with a reliable DPDT switch.
(Perhaps even with an additional parallel switch). When in
normal position, charge the backup battery with a diode from
main battery/alternator. When in backup position the engine
bus is connected to the backup battery and completely isolated
from main battery and alternator (main bus). That will survive
a main battery short and also alternator over-voltage event
(with a OVP device installed).
Finn