reciprocating piston circuit 2

[subsearobot]

Hi experts!
I am designing a 'dumb' system (no external sensing or control signals) for an Antarctic core sampler. I need to raise and drop a weight repeatedly to hammer the core barrel into the sediment. This needs to happen under 2000' of ice.

I think I have seen a circuit (too long ago to remember) that causes a ram simply to cycle back and forth. basically, when the piston hits the end stop, driving pressure climbs, and this pressure is used to reverse plumbing with some sort of logic valve. No controller necessary.

Is this possible? can anyone point me to literature? I have googled "reciprocating" type searches with no avail. Sounds like simple automation, but I do not quite grasp it...

cheers!

 

[hydtools]

Use a motor/roller chain setup with a pickup on the chain to raise the weight. The weight then free falls to impact.
Google Cyclone breakers. It's simple and uncomplicated. The motor runs constantly, no starting and stopping. I am assuming the work in vertical so that gravity drop will work.

A problem with the cycling ram is if you don't allow dwell time at the point of impact, production will be very low or not at all. Without dwell, impact transfer to the core barrel is low.

Ted

 

[EdDanzer]

Wandfluh makes a valve that switches cylinder direction after an adjustable preset pressure is reached. We have built a prototype well casing driving hammer using this valve and a hydraulic cylinder. It works very well if you vibration isolate the valve from the frame. This valve will provide enough dwell / hold time at the end of the sroke to tranfer the force better than just dropping a weight free fall. When a weight drops free fall in a casing hammer it bounces atleast once.

Ed Danzer

http://www.danzcoinc.com

http://www.dehyds.com

 

[hydtools]

True, the hammer may bounce depending on how much initial impact is absorbed by the casing being driven. The hammer will have settled before the lift picks it up and there is no rebound reaction into the supporting structure.

The hydraulic cylinder will experience a lift in the time interval between when the casing is impacted and the valve responds to return the piston. The structure supporting the cylinder must be capable of handling this hydraulic lift.

Ted

 

[PNachtwey]

A problem with the cycling ram is if you don't allow dwell time at the point of impact, production will be very low or not at all. Without dwell, impact transfer to the core barrel is low.

I don't understand this. It is the kinetic energy that does the real work ( force*distance). After the cylinder stops the kinetic energy is dissipated. Dwell time doesn't do anything unless there is enough force to simply push through in which case the the lifting part is required anyway.


Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[hydtools]

Impact energy transfer is not instantaneous. The time to transfer is small but finite. If the piston is returned too soon because of an early signal to return, energy will be absorbed by or returned to the hydraulic system and less will be transferred to the casing. Delaying piston return too long after impact wastes time in the whole cycle with no benefit to the delivery of impact energy. The dwell I am referring to is that short period of time from just before impact to just after impact to deliver maximum energy to drive the casing. Just experience designing hydraulic breakers which are sometimes used to drive posts and pipes.

Ted

 

[subsearobot]

Ted,
dwell time is not too large of an issue in this system for a couple reasons. first, the entire package (hammer, HPU, control electronics etc) must fit down a ~500m long hole in the Antarctic ice that is 8" in diameter. My pump must be pretty small. second, I have 5 shaft HP available. this of course dictates flow rate. cycle times will be at a best 3 times that of the free-fall time for the mass.

I was looking at a pilot operated 4-2 valve from hydraforce. I would sense the pressure developed at the end-of stroke. pressure to raise the load will be less than full system pressure. This system has me a bit concerned, however. I think there could be false readings such as when accelerating the mass upward at the start of the stroke. I like Ed's suggestion if I stay with the piston solution. I am also looking into the feasibility of the chain system like you suggested. And for due-diligence, I am looking at using an ACME shaft and a triggered half-nut. All are fun solutions, I think simplicity will take the cake.

ED, I will look into the wandfluh valves- I used Wandfluh for another project recently, and was very happy with both their function, and the level of support coming form the US division. Using a time delay eliminates false pressure readings.

thanks all!

 

[hydtools]

Consider using regenerative flow on the down stroke. Connecting the rod end to the head end flow. This will give you higher velocity for given pump flow rate since the effective area will be that of the rod area which is smaller than the piston area.

Ted

 

[EdDanzer]

When I did the design for the cylinder driven casing hammer the goal was 2G of acceleration in a relative short (10”) travel. The hammer weighs 600 lbs. The cylinder has a very large rod to bore ratio and required a special piston seal to keep it from shearing off when the hammer hit the anvil. The flow rate will determine the maximum travel speed of hammer, the bore and rod size of the cylinder relative to the flow and pressure will determine the rate of acceleration. If you do not need higher number of hits per minute or high acceleration rates the rod can be smaller relative to the bore than what we did. Because of the high retract speed we use a pilot operated check valve to dump additional flow from the butt end of the cylinder. It is important to have a large enough return hose to keep the back pressure down for proper function.

Subsearobot, I doubt you can stand 3G’s (three times that of free fall). The mass that anchors the cylinder will have to be 3.1+ times the weight of the hammer to generate 3G’s of acceleration otherwise the cylinder will move not the hammer. Once the hammer strikes the anvil it will start moving, soon after the anchor mass will have to start moving down with the anvil. We tried 1/8” of free travel between the anvil and the anchor mass and had to increase it ¾” to make it perform better.

Ed Danzer

http://www.danzcoinc.com

http://www.dehyds.com

 

[PNachtwey]

Consider using regenerative flow on the down stroke. Connecting the rod end to the head end flow. This will give you higher velocity for given pump flow rate since the effective area will be that of the rod area which is smaller than the piston area.

That limits the ability to accelerate because the effective area on blind side is now just the rod area. This may be OK IF the rod is big but then there isn't much gain either.


Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[hydtools]

Peter, the smaller area just increases the pressure requirement. We don't know the system pressure capacity. That will be up to subsearobot's design. So far we just have generalities without specifics.

Ted

 

[subsearobot]

Ted, I'd heard of regeneration circuits, but never had the requirement. This seems like the perfect opportunity. The piston (mounted rod-down) will release the weight at the top of the stroke, then drive down to retrieve the weight. the downward stroke is a high flow, low pressure requirement. Once the weight is latched, I need the pressure to accelerate the load upward. coupled with my limited size requirement, which dictates a low-displacement pump, this will help tremendously.

System pressure may be quite low, dependent on ram choice. We typically use Bimba 500s (500 psi). they are quite cost effective, and *relatively* corrosion acceptable (cheap enough to change out at service intervals). At this low pressure, I do worry about parasitic pressure losses due to my high flow requirements. I still need to quantify.

This high flow req. brings me to another question, but I will start a new thread for that...

cheers all

 

[EdDanzer]

If you are trying to get much acceleration out of a cylinder you cannot use regeneration with the valve I describe as the force required to accelerate the load will not allow much back pressure.

Actually you will use more pressure to accelerate the load in a short distance to even reach 1G of acceleration than you may realize. If you do a force and velocity calculation for your hammer weight you will see how difficult the application is.

I doubt an off shelf cylinder will provide the life or performance you need to make a cylinder operated hammer be effective.

Ed Danzer

http://www.danzcoinc.com

http://www.dehyds.com

 

[PNachtwey]

Ed , don't those hammer systems rely on accumulators? Where would you put them in such a small place? Hydtools first suggestion makes sense but the limit would be just the energy released when the hammer falls due to gravity.
I like the simplicity. The big question then is how far must the mass be raised and how big must the mass be to impart the necessary energy on the ice.

Reaching 1g is easy if you just let the weight fall using Hydtools' idea. Getting more than 1g is going to be more expensive.

What is really required is an estimate of how much energy the hammer has at each stroke so we can tell if 1g acceleration is enough or if more is required. Knowing the mass that is accelerated is important too.



Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[subsearobot]

Peter,
This is the root of the question, for sure.

I have inherited this project form another engineer. When I began, I checked many details, but one that I took for granted was the mass and free fall height. Now I am checking this, intuitively, something is not correct.

I have based the design around 100# mass free falling (1g acceleration) a distance of 1 meter (~40"). The impact energy is significant. (based on a 5" sched 80 core barrel, impact force is ~84000#)

The client (yesterday (!)) told me that the maximum shear strength of the sediment is 200kPa (29 psi). I am still looking at how to equate penetration depth/ blow to this low shear number. Now, I need some civil engineering advise! I did a back of envelope calc, and if it is correct (which it must not be!) this impact will drive the core tube 160" into the sediment (neglecting friction etc).

cheers

 

[EdDanzer]

The hammer system we developed does not have an accumulator. The 600 lbs hammer has a 1 ½” bore cylinder with a 1 1/8” rod and 10” of hammer movement with 16 gpm feeding it from a gear pump. The Wandfluh valve and the PO check are the only parts besides the cylinder required to make the system work. The total valve and cylinder package could be configured to fit in an 8” circle if length is not an issue. The well drilling casing hammer requires a very short complete system length.

It will take more than 40” of travel to reach 1G of energy if you are not providing acceleration force.

Ed Danzer

http://www.danzcoinc.com

http://www.dehyds.com

 

[hydtools]

The impact energy is 333 ft-lb. W*h = 100*40/12
How did you arrive at 84000 lbs impact force?

Penetration will depend on friction resisting the core barrel movement. Friction force x penetration distance = work done against the friction force = impact energy. Work = energy

Ted

 

[PNachtwey]

How did you compute 160 inches? You know that can't be right.

I am assuming the mass is free falling as in Hydtools' method.

If the pipe is very sharp then shear force 8*?*29 or about 729 lbf/in. 8*? is the circumference of the pipe and the length of the shear distance. This distance is multiplied by the shear strength to get the force required as a energy as a function of distance. Think of this like a spring constant where work or energy=(1/2)*K*x^2, the integral of f(x) dx. The energy when contact is made is 100lb*1g*1m=3927in*lbf. This is simple energy=m*g*h. ( Mathcad works out the units for me). This is equal to (1/2)*729*(lbf/in)*(?x*in)^2. ?x=3.28in which I think is very optimistic because no other friction is taken into account and the shear strength seems low.

Another trick is to vibrate the the machine. It makes the ground liquify like an earthquake and softer. Perhaps some motor with an eccentric weight can help but at 2000 ft down it would be bad if the machine vibrated apart.



Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[subsearobot]

I found the impact force by assuming the pipe acts like a spring. 1/2 kx^2=mgh. solved for x, then plugged x into f=kx. this assumes pipe sitting on bedrock. I did this calculation for a buckling analysis. I did, however use this force in the shear equation where 160" was the result. I realize this is erroneous, because once the tube moves, impact force is significantly reduced.

Pete, this makes perfect sense, thanks. work in energy, not force. I have some frictional values, so I can refine a bit.

 

[Gkranz]

Subsearobot,

You are getting some incorrect information in these posts. If you want to know how to do this job, go to American Piledriving Equipment, Inc web site and get some good information. You are basically driving a pile into the ground; this site will have good information on how to do this.

I design all their impact hammers and just finished the X13 impact hammer rated at 12,000 lbs and blow counts exceeding 38. Other than the size, what you are trying to do is pretty easy. Also, forget an off the shelf cylinder, never will hold up. Valves and cylinder are always custom designed for application. Good Luck.


Best Regards,
Westerndynamics.com