OP, this is an install issue, not a Jeep issue.
Jon, those are not screws, they are bolts.
You install a bolt the way the MFG recommends you install it.
If it is meant to be installed dry, you install it dry as a bone, and no other way. If they say put Loctite on it, put Loctite on it; oil, then do oil; RTV the threads, then RTV it up.
That being said, there is a large variance on what folks do with water pump bolts that may or may not enter the water jacket. Many folks put a coating of RTV on it, and put it in. I prefer using Permatex #1 on water pump gasket and bolts. If there is a doubt that it enters the jacket, I will RTV or Permatex by application.
I suspect the aforesaid fasteners were not installed to spec. When you install anything held by more than one fastener, tighten all around, then go back and make sure all fasteners tight. Sometimes gasket material compresses during the torquing process and if you don't make sure they are all tight again, you may find yourself looking at loose fasteners.
A "screw" is a threaded fastener that mates with the part it is installed into - be it machine thread, tapping, wood screw, or whatever.
A "bolt" mates with a separate threaded part of some variety - hex nut, wing nut, knurled thumb nut, or whatever.
Ergo, them is screws. Proper catalogues refer to them as "hex head capscrews" - in that particular case, they would be "flanged hex head capscrews," due to the flange under the head (that is larger than the heat itself.)
Yeah, I'm being pedantic. It's just who I is!:helpme:
The "clean, dry" spec itself is of only limited use - how many of us make sure that screws are "clean and dry" before they go back in? (Yes, this also means cleaning out the hole it goes into!)
However, there are published values for the modification of the "clean, dry' torque value as given in the manuals, depending on the compound added to the threads. That's why the manual values are specified to be "clean, dry" in the first place - because as soon as you add a lubricant, you have to change the value (unless it's formulated to not provide any actual lubrication.)
The idea of measuring installation torque is to provide an indirect measurement of tensile preload in the given fastener - the act of lubricating the threads means that less torque (turning force) is required to achieve the same installed tensile preload ("stretch.")
It should also be noted that, when a lubricant is added to the threads, the "error" for the preload (how much you get, vice how much you're supposed to get) decreases as well. I really need to find my notes on the project I did for Materials & Processes in college a few years back, they're illustrative.
As I recall (using ten samples of each value. 1/4"-20x1" hex head capscrew, black oxide, threaded into 10L20 mild steel with cut threads, no hardening. The same torque wrench - a beam-type - was used for all fasteners. I think I torqued everything to 25 pound-feet.)
- "clean, dry" got me something like 20-25% variation in tensile preload for a given torque
- Using LocTite got the same tensile preload for applied torque, but with only 18-20% variation (lubrication effect was negligible)
- RTV had similar results to LocTite
- Clean water ran something like a 5% reduction in torque to get similar preload, variation was 15-18% (yes, water lubricates!)
- PTFE ("Teflon") paste got similar preload with 10-15% less applied torque, and with 12-15% variation
- Engine oil or chassis grease achieved the same tensile preload with 20-25%
less applied torque, and with a variation of 15-18%
- Engine assembly lubricant yielded similar preload, 30-35%
less applied torque, variation of 10%
- Anti-seize compound ("never-seez" - I checked both Ni-base and Cu-base) yielded the most dramatic results: similar preload with 50% applied torque, variation <5%!
Deforming a single point in the threads (enough deformation to be visible) was effectively negligible for torque/tensile preload relations, but deformed threads are great for reducing loosening due to vibration.
No, I didn't conduct this as a scientific study - but all of the screws were from the same lot, and I cut all of the internal threads and checked them with a good plug gage before I went to the next stage. Doing this experiment verified what I'd been taught years ago:
Use lubricants specified in book. If none specified:
"clean, dry"? 100% of "book value"
LocTite or RTV? No change.
PTFE paste? 90% of book value (I didn't test PTFE tape, I don't use the stuff.)
Engine oil or chassis grease? 75-80% of book value
Engine Assembly lube? 65-70% of book value
Never-seez?
Half of book value.
If deforming threads, don't get stupid. Single point for external threads, single thread for internal threads. (N.B. Deformed-thread locknuts are used throughout the suspension of most vehicles, and those torque values are still listed as "clean, dry." Lubricant effects are similar.)
I've been doing it this way for as long as I can remember, I've never had it cause me any trouble.
Newer vehicles are moving toward a "torque angle" method - I've been using that in plumbing, hydraulics, and pneumatics for years. If you have a specified torque angle, lubricants become moot (since preload then becomes determined by
how much you turn the fastener, not
how hard.)
What I do may not be approved in the manual, but I've been doing it long enough that I'd know if it was going to cause trouble by now!
It should be borne in mind that what is built is done for the convenience of the builder, not the mechanic. (That's why Torx screws were invented.) The mechanic - especially the hobbyist! - is likely to want to make sure his job isn't going to come undone, or is going to stay together until he needs to take it apart again. That's why thread compound producers and aftermarket fastener companies give with these rules - and, having experimentally verified them, I'm inclined to agree.