There are generally two points of view:
a. The engineers who designed it know best -- RTFM
b. Drive it like you stole it
Both points of view have their merits, but lets take a moment to look at what the break-in is for, and why the manufactures make specific recommendations.
Before I discribe how I broke in my car, a few comments on how and why.
A new engine (or any other mechanical device) is a little like a new pair of shoes. Everyone knows from experience what would happen if you put on a brand new pair of hiking boots and went for a long run. Big time blisters on your feet. In a mechanical device the same sort of high pressure injury occurs, but an engine or a gear does not blister, and cannot heal. You either break it in correctly or you damage it permanently.
The Break in process is actually a carefully controlled form of wear, like the new shoes, it is a chance for the mating parts to "adjust to each other". In a shoe and your foot, most of the adjustment takes place as the shoe leather stretches and softens from repeated flexing. To a lesser extent your feet will form a callous to protect any area where a pressure point persists.
No engine, bearing, gear or other mechanical device is perfect. If you look at a machined surface under magnification a piece of steel that looks to be smooth and polished is actually as rough as sand paper, only the peaks and valleys are on a very small level. When a machinist turns a shaft to a certain size, he is actually working to an acceptable error called tolerance. The crank shaft main journal may be specified to be 2.5 inches + or - 0.0002 inches and a surface roughness of maybe 50 millionths of an inch. A fine human hair is about 2 thousands of an inch in dia. So the acceptable error in diameter of the bearing is plus or minus 1/10 of the diameter of a human hair, and the acceptable roughness is about 1/40th of the diameter of a human hair.
The bearing clearance is the space between the bearing journal and the bearing shell surface where the oil film is supposed to exist. Lets suppose the clearance on this crankshaft is supposed to be 5/10000 of an inch to 7/10000 of an inch. Now suppose you take a bearing that is on the small side with a nominal clearance of 5 ten thousandths and put it on a journal that is at the large limit of 2.5002 inches. You have an actual clearance between those two high points of only 3 ten thousandths of an inch. On another journal in the same engine you might have a large bearing that has a nominal clearance of 7 ten thousands of an inch and a small journal that is only 2.4998 inches in diameter. On this bearing the actual clearance would be 9 ten thousands of an inch. Needless to say the bearing that has a true running clearance of 3 ten thousandths of an inch is going to have much less room to expand if it gets hot than the looser bearing journal that is at 3x the clearance.
In real life the crank shaft journal actually moves around inside the bearing. On average the bearing maintains a thin oil film between the moving parts but from time to time, under high loads or just due to random vibrations and variations in oil pressure and such, the bearing and journal will move into very close contact. Over time this will "burnish" the high spots down so the average surface is much smoother than it was when it came out of the factory. If you imagine the sandpaper like surface it had from the factory on a microscopic level, over time the appearance of the surface roughness will begin to look like a lot of table top mountains, as the upper most points on the the surface get either worn or beaten down by millions upon millions of small impacts.
The break in process is intended to facilitate this "wearing in" process with out giving your engine the mechanical equivalent of a blister. [ what is a mechanical blister ?? on a fly wheel or a brake disk it is a premanent hard spot created by overheating a small high spot which is rapidly cooled by the metal surrounding it. This portion of the flywheel or brake disk will never wear at the same rate as the rest of the disk and will cause chatter and uneven friction for the life of the part. On a bearing it can be a hot spot on the surface that either hardens or softens the bearing surface or destroys the thin wear resistant coating put on by some manufactures. On a gear tooth it will be change in the hardness of the surface of the gear tooth that will make it wear or break under load. It could also lead to micro-cracking due to the rapid heating which will permanently weaken the gear. On a piston ring it can be scratching or scoring of the cylinder wall so the engine will always use oil to a softening of the piston rings due to over heating that results in poor ring seal and low compression and oil consumption.]
How do you allow controlled wear? You have to create a situation where you have sufficient load to create the brief periods of close contact required, without over heating or damaging the bearing surfaces. The problem is that each mechanical component needs slightly different conditions to allow this to happen.
The RTFM approach to auto break-in will give you a reliable car that does not require any warrantee work due to problems you created by mistreating the car in its first few miles of use. IT DOES NOT necessarily give you the best running car, or the most horsepower from your engine.
The drive it like you stole it approach will MOST of the time, give you a powerful engine, but may lead to various trips to the dealer for things you pushed too hard too soon during the first few miles of driving.
The best break in approach is about half way between these two extremes. The cam shaft and the cam followers are among the most heavily loaded parts on an engine, and need special care during the first hour or so of engine operation to avoid permanent damage. The bearings through out the engine and the gears in the transmission and differentials also need a pattern of alternating load and time to cool starting at low intensity and gradually increasing stress, to avoid permanent changes in gear hardness. The piston rings require brief periods of heavy load (hard acceleration), to seat the rings, followed by time to cool. The brake pads and brake disks as well as the flywheel and clutch need a progressive pattern of bedding in for the friction material and friction surface to get properly conditioned to avoid permanent damage due to over heating or poor performance due to improper bedding.
Most people think of the break-in process as applying to the engine, but it actually applies to the entire car and all its systems.
So what do I recommend for a break-in procedure?
First, cater to the cam shaft and the bearings, piston rings and cylinder walls in the engine. Start the car and hold the engine at a slightly fast idle (about 1500 -2000 rpm), let the engine run at this fast idle with slight variations is speed for about 30 minutes or until the engine is at full operating temperature for a minimum of 10 minutes.
Second, begin the bedding in process for the clutch, brakes, and the break-in process for the gears and piston rings/cylinder walls.
Take the car for a short drive (drive it home). Make a point to not slip the clutch too much, or ride the brakes too much, accelerate briskly up to the recommended rpm limit the manufacture gives in the break in guidelines, but be sure to give plenty of time between the periods of brisk acceleration for the engine to cool any hot spots. After you have several minutes of driving on the car, in a safe location, make some moderately hard stops to begin the bedding process on the brakes, but space them well apart so you never smell hot brakes. By now you should have a feel for the clutch, make a couple moderately hard accelerations which load the clutch a bit, again with plenty of time to cool. By now you should be home. Park the car someplace where you will be able to see any drips, and let it cool completely.
After the car has cooled down, check all your fluids to be sure all your levels are good. Look under the car and check for drips.
Take the time to give the car a good visual inspection. Think like a pilot giving a plane a pre-flight. Grab stuff and try to shake it, check tire pressures, oil levels, get out the owners manual and learn where the transmission fill is, where the brake master cylinder is, where the fuse boxes are, how to turn on the 4 way flashers etc.
Now your ready for phase 2. Take the car for another drive, where you have a lot of stop and go driving. Find a residential neighborhood that has lots of stop signs and learn the feel of the clutch. When conditions are safe, make a few brisk accelerations to seat the rings and bed in the clutch, then pull your speed down briskly (about 50% effort) with the brakes to help bed in the brake pads. Again be sure to give the car plenty of time to cool with easy "grandma driving" between these bursts. (be careful your brakes will not grab at full strength until the bedding process is complete - don't put your self in an emergency must stop situation !!)
Find a low traffic highway and make a max effort acceleration up the on ramp to highway speed, revving the engine to the max rpm listed in the manufactures guide lines in each gear (lets not go 140 mph here guys just 1-4 th gear ). cruise for a mile or two, to let everything cool and then find a safe place to make a 75% effort stop or two. Do this several times then take it home and let the car cool over night. Follow this pattern for a couple days, gradually increasing the level of effort on brakes, clutch and acceleration until you have about 200-300 miles on the car. At this point you should be able to make max effort stops and accelerations, but still limit your engine rpm to the max rpm recommendation by the manufacture.
From this point on, drive the car hard (safely!!) and every 100 miles or so, raise your engine rpm limit by 100-200 rpm. By this time you can make long highway trips with no problem, just avoid running on cruise control for extended periods of time. If you do use cruise control, periodically vary both speed setting and/or alternate between 4 th and 5th gear.
Switch to a quality synthetic oil on your first oil change, and by the time your at about 3000 miles you should have finished all the break in that requires special precautions. The engine will continue to gain power for the first 5 - 10 thousand miles as it slowly loosens up.
As an example of how these principles came about, here are some other examples of break in guide lines.
As far as manufactures recommending it, they have, but only to that fraction of their customer base that are likely to understand it. Several manufactures have specifically recommended this sort of high load, with cool down to racers.
The one I'm familiar with is Chrysler Corp when they introduced the Street hemi.
For their racing buyers they told them to follow the following sequence to break in their race engines for drag racing.
Start them up and let them idle at a fast idle 1500 rpm or so for 20 minutes or so. ( most cam shaft manufactures recommend the same for new cam shafts ) Then after checking for leaks etc. let the engine cool down completely.
Next make runs at 1/4 throttle , 1/2 throttle, and 3/4 throttle, each followed by a full cool down. Make a final full throttle pass then go racing.
Carrol Smith one of the most respected racing engineers in the business, goes to great length in his book "Tune to Win", and then bothers to repeat the same advice in a following book about how important this sort of periodic high load, followed by cool down is to proper break in of differential gear sets.
Several major engine builders describe how they dyno break-in an engine and they also use the same pattern. 30 minutes or so of moderate fast idle to warm things up, followed by brief runs up through the engines rpm range under gradually increasing load, followed by cool down cycles.
Mark Donohue the famous race driver describes how he broke-in race engines after rebuilding them in his book "Unfair Advantage". He liked to take the newly rebuilt engine and let it run at a fast idle for an hour or so, with a garden hose running in the cooling system to keep it cool. Then he would put it back in his car and take it to the track. By the time he finished his tuning laps and time trials for qualification, the engine was ready to race.
Graham Bell author of the book "Four Stroke Performance Tuning" gives an example of a dyno run in process he uses on all his race and rally engines. To make it universal I converted the numbers to % of maximum torque.
rpm Torque min
3500 25% 10
4000 33% 30
4500 45% 30
5000 56% 30
5500 66% 30
Followed by cool down, adjust tappets, and retorque heads then full power runs for tuning purposes.
Kudos to Larry for sharing this knowledge