Yeah, you mean like San Francisco or something like that? Oh, if I could go back and educate young Scottmitchell74...
So I've been doing a few experiments... 1) Going the other way, so it's climbing the 11" curve instead of 9.75. The Kato was able to pull 7 cars, but barely. A lot of slipping. It's in the same spot as the 9.75 curve, so the grade is the same, but any potential track issues are of course not there, or different. Joints are in different places, etc... 2) Going the same way - it can pull 5 cars with the same amount of slipping seen with 7 cars the other direction. So, using the more forgiving curve helps, but then again the other locos I have (LLs) also pull better going up the 11" curve. 3) Clean wheels - wheels are clean. I have a Tidy Track thing and gave all my locos a good cleaning just before the videos the other day. 4) Track issues - I have them, so maybe the Kato and Atlas (maybe being more precision machines?) are less forgiving? However, when not under a load, they all zip around my track as smooth and pretty as anything. I suppose a load will highlight any problems. I wonder if my early successes with Life Like (choosing them was a factor of income and price concerns at the time) have given me a slightly skewed expectation of other brands? Those early-to-mid 90s LL's just happened to be their Golden Years. Because they make up most of my roster and perform so well, maybe I expect performance from other brands to rise proportionally with their prices respective to Life Like?
No one is disputing the heft of the older LifeLike locomotives. I don't own any of them because the first couple I had were very unreliable. I am predominantly Kato with Atlas and IM thrown in. Track laying is as important as equipment too. A transition curve before entering the grade would help immensely. That would allow the the locomotive to ease into the curves and not lose as much forward momentum. I have never worried about the pulling power up grades because I don't go over 2% grade on my layout. And a standard lash-up of locomotives like an ABBA set of F units has no problems pulling the trains up the grades. I also have transition curves and that has helped not just in pulling the train, but it also helps to minimize derailments of the cars at the tail end of a long train.
So now you're saying that,since the motor and gears are turning,their weight doesn't count?So,brass manufacturers only put weight on the towers,because weight anywhere else is of no consequence? Are you saying that if I fill that F7 with lead,it won't work any better? I always thought brass manufacturers were just too cheap to put more lead in,but of course,I'm probably wrong..
The weight of the wheels and gearing do not help with traction. In the auto racing world, most go for light weight wheels, lighter brake calipers, and two piece brake rotors. As it lowers the unsprung weight, and unsprung weight doesn't aid in traction, speed or braking. The weights still count, but being they are spinning, the weight isn't always in the same spot. The flywheels and gears aren't going to be perfectly balanced, but close enough for the tolerances that were designed into it. And being flywheels are spinning, they throw a bit of torque to the frame, its not much, but with it spinning, creating torque, it in theory will cause a bit of lift to the frame on one side. How much, probably so little you won't notice. The general consensus has always been to add weight over the drive wheels. So like on a steamer, you add weight over the drivers, not the lead or trailing truck, as weight there just puts weight down on the rails on wheels that have no effect on traction. But say on an F7, if you add 2 ounces of weight over the fuel tank, in theory, the weight will be spread 1 ounce to each truck, however if you add 2 ounces directly over the lead truck, that weight will all be carried by the lead truck and aid in traction.
You do know I'm a racecar builder,right? This week,I'm negotiating going to work for a 6 time SCCA national champ,I also work for an engine builder,yesterday,I put together a Lotus Elite race engine,and put it in the car.Just for an added bonus,I ran RC pullers in the NRCTPA on national level for 10 years,and won best engineered truck at the NRCTPA World Championship at least twice,and I've built real pullers.I also ran a national championship Motocross team..So,yeah,I have a clue.. The only reason "we" shoot for unsprung weight being as low as possible is because it becomes more difficult to gain traction with high unsprung weight.Unsprung weight is weight that cannot be controlled by the shocks or springs,therefore,any road irregularity has the tendency of lifting the tire off the road,which totally negates traction.The lower the unsprung weight,the faster the upset wheel can react,thus,better traction.In that context,a model locomotive really has no "unsprung weight",unless of course,you're running it 200 MPH,and your trackwork is aweful... As far as rotational mass affecting traction,the actual speed that a model train operates at,combined with the miniscule weights of such objects,means nothing as far as any affect on traction. Now,in your scenario of adding two ounces to the middle,as opposed to over the truck,well,what else could happen??? You put two ounces over the truck,OF COURSE it gonna add two ounces to that truck.Two ounces in the middle? Of course only one ounce is gonna go to each truck..What else would it do??? Now,same locomotive,if you can put two ounces in the middle,and two ounces over the front truck,then you should also be able to put two more ounces over the rear truck.You've added six ounces of weight to the locomotive,and hence,three ounces to each truck.The weight in the middle doesn't magically become less just because it's in the middle,which is what he's saying.Weight is weight,and a model train simply doesn't generate enough speed or rotational force for anything funky to happen..Yeah,it's better to put weight over the trucks,but in a diesel,ALL WEIGHT IS ON THE DRIVERS.Even if it somehow transfers,it's still on the drivers..A 1/4 ounce is 1/4 ounce no matter where,or what it is.The brass industry doesn't JUST put weight on the towers because it's the only place to put it,they do that because they're cheap. And the reason you don't put weight FAR outside the drivers on a steamer is because if you put weight outside,it will cause the drivers on the end opposite the end being pulled to rise up,and loose traction,an effect of imbalance in the chassis..If you had a steamer you only planned to run forward,you could put a pile of weight in the front only,and it would pull your house down.Try to run it backwards,though,it will flip over.But,also put an equal pile in the back,it will increase traction over not putting in weight,but again,reach the limits,it will lose traction because the outer weight will start to react against the weight in the front.A steamer is a teeter totter,no resemblance to a diesel..
Alright! I have scanned through the posts and I don't think anybody has brought this up ... so here goes. I own several lifelike canada locomotives and with out a doubt they will outpull an equivalent atlas locomotive up my 2% grade every time. My comparison engines are a LikeLike GP9rm vs an Atlas GP9 and a LifeLike C424 vs an Atlas C420. These Locos are virtually the exact weight. not only to the Atlas locos slow significantly up the grade but they also speed up significantly going down hill. whereas the LifeLike Locos maintain a fairly even speed, either way. I can only guess that it may have something to do with the gearing.... regardless, to overcome this I pair a lifelike with an atlas and they sort of cancel each other out and the train operates as smooth as silk.
So wanna help me after I will the lotto and slap an LS9 into a Fiero? Figured it would make a decent sleeper...or just a crap car with an awesome engine. But on my car, I really need to lower my unsprung weight...thinking a set of OZ wheels and some Racing Brake two piece rotors.
Anybody that's hung around this forum for any length of time knows this is a particularly irksome topic with me that I've attacked with scales and measuring equipment over the last decade. So I've got my own database of locomotive performance, backed up with measurements. I focus a lot on the difference between wheel materials because two locomotives that weigh the same and are designed the same 'should' be able to pull the same, and don't; it has nothing to do with gearing, motor torque, or anything else - it's purely the adhesion factor between different alloys. Life-Like used/uses a blackened brass/sintered alloy of some kind that's relatively rough and gets great traction, so it's no surprise to me that you're getting better results. Atlas has a very slick alloy on much of their new production and while it might help keep the wheels clean and help DCC, it's horrible for traction. Kato, on the other hand, has used a plated harder brass on old stuff, and some kind of solid alloy that gets a pretty good grip on new production. So the only mystery to me is why your SD40 is performing so poorly, and without the video, I'm still not sure I'd believe it. I've got one and put it back in the test rack. I was curious about what you're saying/seeing on the sharper curves. I noticed that on a 9 3/4 test curve I have that when the trucks are swung over, there's two things happening; one is that the truck wheelbase of the locomotive is so tight in that curve the wheels are beginning to lift slightly, the second is that there's not enough vertical freedom to keep the trucks on the rails if there is any, and I mean ANY variance in the curve surface vertically, the frame doesn't let them tip at all. It will lift wheels off the rail. Now you'd think the weight would just transfer over to another wheel that is in contact. But it's an observation, it's right on the bleeding edge of what that loco can do. With what you reported on the 11", I'm still wondering if you've got a slipping universal on a truck, and the only way you can prove that is to get right down and eyeball the thing, not necessarily on that curve, but on any piece of track, and watch the wheels start to slip as you apply power. Hold down a coupled car behind it, let it slip, and watch the front and rear trucks to make sure they are actually slipping in unison. It's a long shot, but it still should be doing better (at least on 11") than what you're seeing. I rate a Kato SD40 chassis at about 12 cars on 11" radius on a 2.5% grade on my layout. It's right on the edge, but I can do 25 cars up that hill with two six-axle Katos. In general, my Katos are right up there with the Life-Likes, leaving the Atlas way behind. But I have had problems with nearly all Life-Likes with building up 'dirt' on the wheeltreads that has to be polished off with running; if they've sat around for a few weeks they'll stall until they've got some fresh run time on them. It was bad enough with my LL SW8 that I swapped mechanisms with a Kato NW2, end of problem, and no loss of pulling power either (measured with scales and the drawbar puller I built). Road units (like the GP20) are usually MU'd, so it's not an issue, but it sure was a PITA with a single-unit switcher creeping in a yard.
It is amazing how even the more broad curves can affect a locomotive's performance. After my initial reading of this topic, I observed closely, one of my Atlas eight wheelers as it went into a Kato thirteen and three quarter curve at twenty SMPH. That curve, you would think, at least, should be more than generous for so small a locomotive. Still, it slowed to about fifteen SMPH. When it came to a one per cent grade, I had to apply a little more throttle, as it appeared that it was about to cog. It was pulling three Athearn thirty six foot cars and the Athearn wood caboose. Two of the cars and the caboose were on Athearn trucks, one of the cars had MT.
The reason performance drops off in corners is because there are no differentials on locomotive & rolling stock wheels.Since the wheels on the rolling stock need to be going two different speeds because the inside track is shorter.The tighter the turn,the more drag is created.Different scenario on the locomotives,the tighter turns actually break the wheels loose when it isn't pulling anything.Where I live,the track has a LOT of tight curves and hills.On real trains I've been on,the engineer has to actually pull the train DOWN hills with the locomotive if he was going through a series of tight turns..
The 'built-in differential' on railroad wheels (and to a lesser extent, the models) is a properly tapered wheel tread and flange fillet. As the wheel goes to the outside of the curve, the wheel gets slightly larger due to wheel tread on the outside, and smaller on the inside, and assists in keeping the wheels from slipping at all. When you hear flanges and wheel treads screeching, that's when it's actually sliding and not keeping up. That, in conjunction with track speed and superelevation, is supposed to keep the cars riding down the middle of a curve; as a track inspector you're always looking at the high and low rail for excessive headwear indicating that speed or superelevation is set wrong; when it's set "just right" you won't see flange shine the upper OR lower rails, but that's the minority of times. On a lot of ex-65mph track on E-L that had been downgraded to 40mph or less, the low-inside rail was just getting destroyed and the high rail wasn't touched because the superelevation was WAY too high to be running that slow. I've seen beautiful multiple reverse-curve track on BN that was PERFECTLY set, and talked to crew that said, yup, 'you hit that curve a 42mph, period, and listen, and it won't make a sound'. I've seen this on some models; namely my old Kato F's. The wheel treads are too flat on those. and you can actually hear them slipping slightly on curves. That's the only ones I've ever seen do that, and wow, is it evident. Most models have enough taper that they 'sort of' work, you don't realize its working as well as it does unless you manage to see truly flat wheeltreads. One of the techniques that 'old heads' will use coming downhill is to take a single steady air set on the train and apply some locomotive power to keep it stretched out; doesn't apply to dynamic brakes, but it's a time-honored method called power-braking and has been outlawed by many railroads as it can really beat up brake shoes when used excessively. The physics of a downhill vs. uphill train on curves and engineer braking patterns and methods take most of the 'perfect science' of track and wheel design and throws it out the window. One of the odd things that happens with tapered wheels is that if you are running through curves too slowly, the drag can actually increase because there's not enough centrifugal force to move the cars 'up' on the curve where they belong, the cars are 'rolling low' and now the wheel taper is actually working against you. If you've ever been on a passenger train on a main line and got a restricted signal, and roll through a sharp curve at low speed, you'll hear and feel all kinds of odd things going on that normally aren't there as the 'design speed' of the curve is no longer effective.
Another thing to consider is that Atlas and Kato use an outside bearing, pimpoint, pickup. This constrains the sideplay of the wheelsets so they don't conform to the curve and may add more drag. An inside bearing setup like a Lifelike E8 should have more sideplay and should allow the axles to move independently to conform to the curve. Adrian
That is true for the prototype. But do our models have this fillet and if they do is it sufficient to work on sharp curves? I don'tthink so. One can hear the flanges on cars squeal as they traverse Horseshoe Curve on the old PRR mainline near Altoona. That curve is a compound curve [two different radii] which in N scale would be about 45-48 inch radii. But few of us have such generous curves. Even the majority of Ntrak corners have a maximum of 39-40 inch curves [four foot corner module]. This also points out one other problem regarding engine wheelsets and also any metal wheels on rolling stock. Because they need to be insulated from each other they must be made in three pieces. If one wheel must slide in a curve that would place a twisting force on the wheelset and maybe affect the gauge of the wheelset.
I consider Horseshoe one of those places that you can't win; you've got all kinds of train dynamics going on with curves, grades, pushers, conventional braking and dynamics, and pretty much whatever you do, at least some of the train is not in equilibrium as on a flat curve at speed. I've been on the curve and actually seen a car being pushed uphill with all eight wheels sliding, and had to flag the helper crew to stop it. The physics and forces there are just incredible when you can SLIDE A CAR UPHILL and not even know it from the cabs. I'll still subscribe to the opinion that the tapered wheeltreads on every 'good' N wheel serve a purpose, and they actually lower curve drag. The effectiveness of a model fillet has been argued for years, other than MT wheels, a lot don't even have it. Old NMRA studies said it was effective, I don't know, should be, but the physics for a model are way different. My dynamometer car with the spring scale inside of it gives me a pretty good feel for curve drag, and how I've had to compensate grades for the curves to see if I could keep total train drag relatively constant. The very biggest impact I've seen on model curve drag is dirty, gunked-up car wheels; I don't know if it's because the taper is gone (outside edges tend to build up first with the inside edges building up last), the tread can't 'slide', or both, but I can knock down curve drag as much as 20% with an aggressive wheel cleaning program; at least that one I actually measure for myself. This is one of those 'angels on the head of a pin' debates for sure, and it crosses multiple scales - the HO guys argue about RP25 wheel contours even more than we do. But one thing for sure is that dirty car wheels create a LOT more curve drag than clean ones even if I can't explain exactly why. But if you go back to the original videos he's got, the SD40 train is slowing as it is 2/3 through the curve.... so it can be a multitude of drag and locomotive factors all at work here.
