95 Civic overheating

We have a 95 Civic (125K miles; very clean and in pretty good shape)
which we use occasionally. About 1.5 years back the temp indicator
rose to almost 75% (had never happened before). I opened the hood and
the coolant reservoir was empty. I topped it and took it to the local
Pep Boys for a check. They said that the radiator was leaking and
replaced it.

Now, 6 months back I got a high temperature indicator and once again
the reservoir was empty so I filled it but then every few weeks the
tank would go empty. I examined closely and it seemed like the top
hose connecting the radiator to the engine might have been loose (had
some leak around it) so I fixed it.

About 2 weeks back when I took the car for an oil change I asked the
mechanic to see if there was any evidence of a leak around/beneath the
engine but he didnt find anything.

Now 4 days back once again the temperature rose to almost 90% (just
below the red line) on a hot day as I was nearing home. But this time
reservoir was full, however I opened the radiator cap and directly
filled it with coolant (took almost 0.25 liters in which was puzzling)
and the car has been running fine since then.

However I still took it to a local mechanic (he's pretty good) who did
a compression check on the radiator and it seemed like it was slowly
leaking. He also did a combustion leak test (http://
www.arrowheadradiator.com/head_gasket_or_combustion_leak_test.htm) and
found that there are exhaust gases in the cooling system (the blue
liquid instantaneously changed to yellow). He thinks its the head
gasket and quoted me $900 for the job (including machining the head).
Now before I shell out $900 I want to make sure that it indeed is the
head gasket. Some questions which I have are:

(i) Could it also be a leaking water pump (hasnt been changed in a
while even when the timing belt was)?

(ii) Is the combustion leak test perfectly accurate (this morning I
ran the engine for about 10 mins and didnt observe any bubbles at all
in the radiator).

Any advice from you talented folks on rec.autos.makes.honda is much
appreciated.

 

if the combustion leak test shows positive, it's the head gasket. end
of story. it tends to happen if the motor gets overheated, and a
defective radiator will do that.

as for replacement, do the math on the labor, and be shy of head
machining. shops do it from habit, tradition with poor quality detroit
heads that used to warp, and because they don't need to spend
unprofitable time cleaning very carefully. the result however is that
the surface finish is seldom sufficient to re-seal long term and it will
blow again before too long. in addition, the cam timing will now be
slightly retarded - approx 2 degrees per mm for the civic d-series
engine. added to that is a lot of shops use abrasives when cleaning the
block. this then resides in the cylinders causing increased wear,
compression loss and oil consumption. all in all, a bad deal unless you
find an expert that really knows their stuff.

the other option is replacement. pulling and replacing the whole engine
is relatively quick. for $300-400, you can buy a replacement jdm engine
and avoid all the above, and its labor charges. again, do the math and
figure out what you want. and if you plan on keeping the car long term.

 

Update:

I took the car to the local Honda dealer who diagnosed the problem to
a leaky water pump. Now I am really confused as my mechanic thinks its
the head gasket (based on the compression leak test; see my original
post) but Honda dealer says otherwise.

My mechanic can do the head gasket as well as the water pump timing
belt for ~$900 (I live in the mid west) so I am thinking of replacing
both. Does changing the head gasket has any adverse effect on the
engine performance because t works really well right now (just as
new) ?

 

ask them whether they did the chemical test!

if the head gasket diagnosis is incorrect, it's foolish to replace it.
if it is though, go ahead and do the whole lot. beware of the issues
cited in my previous response.

 

| www.arrowheadradiator.com/head_gasket_or_combustion_leak_test.htm) and
| found that there are exhaust gases in the cooling system (the blue
| liquid instantaneously changed to yellow).

This liquid costs US$50 ! 1 can use ordinary pH indicator
strips ( has 4 colours ) to detect a rise in acidity ( caused
by CO2 dissolved in coolant ).