Help on latch system design 2

[huangcunbing]

Hi guys,

Does anybody know or have experience with latch system design, especially for mechanical latch mechanisms?

I am troubled with the problem that I have to redesign the system to adjust the desired mechanical release forces. I have been trying to locate the references specially focused on various latch systme design and calculations. but I have not found it.

If any clue or advice about this, it shall be appreciated.

Thanks

 

[MikeHalloran]

One way to start is to analyze the system that you have, and see if your math model comes anywhere near the forces that you can go out and measure on an actual mechanism.

So the first step is to measure what you've got, and the second step is to model it, and the third step is to revise the model until it "Predicts" what you have measured. Then the fourth step is revise the model until it predicts what you want, and the fifth step is to build that unit and test it.

The model will probably have to go into enough detail to compute the spring rate of many of the elements, especially if there are no springs per se. It may have to include clearances, i.e., backlash, as well, but I'd start with a simple model.

If none of that makes sense to you, you'll need to rent an engineer.



Mike Halloran
Pembroke Pines, FL, USA

 

[huangcunbing]

Thank you Mike,

Actually, I am a mechanical engineer at a drilling tool company, but I am not familiar with the latch design.

The latch system is just one part mechanism of the the product I am working on. That is why I have to understand the latch system in hope of finding out some design formulas or tips to revise the product.

Thank you again for your valuable inputs.

Michael

 

[EnergyMix]

huangcunbing

A couple starting points: Find a machinery's handbook and look up latch mechanisms. Do a search on here on latchs. There have been quite a number of threads on different types of latches and different problems people have had with latches. One of them may "resonate" with your problem.

Right now it seems you're not quite sure where to start -- but we're not quite sure where to point you as it's not as simple as one formula solves all.

 

[zekeman]

Unfortunately,the vagueness of the inquiry is why you are not gttting a good response.
Without revealing the mechanism, is it possible for you to show us the existing latch that you have and then maybe someone can give you a clue as to where to begin?

 

[huangcunbing]

I appreciate every input. Thanks a lot.
What am I working on?
Here I attach the right cross section picture showing the latch mechanism. The lathes (red) with different segments circumferentially site insides the grooves and are hold tightly by the disk spring preload with the back rings (the two small pink ones) which are between springs and the latches. There is a middle ring (big pink one) between the two grooves and hold the latches with the inside angle.
Axial force is applied to move the mandrel in order to release the latch to cause firing. The force is affected by the spring preload force (force to hold the latches), holding angle (contact angel between latches and rings) and the groove angles.
I have conducted static force analysis and come up the relationship between preloading spring force and the axial force applied on the mandrel to release the latches. I try to predict how much forces I need to release the latch system from either direction. However, the calculation works well with some configurations, and some need modification.
Actually, it is a mechanical latch system for our drilling jars. Due to the limited knowledge and close references availability, I am not so certain that my analysis works well and need verification. That is why I come here for your guys’ input. I am newcomer to the industry.

What help I need from you:
1) Any comments on the latch mechanism.
2) Any references to help me to understand it better.

Many thanks!

 

[jlnsol]

I printed your drawing and try to help you. Hope to post something tomorrow.

 

[huangcunbing]

Many thanks, jlnsol! I look forward to hear your advice.

 

[jlnsol]

Since I am not familiar with this type of latches I try to take it apart hereunder and find out how it is intended to work. Maybe we can make some progress then.
In the attached drawing I added balloons with letters to indicate each relevant part. The only latches that I know of are quick disconnects for hydraulic lines. The actual latches in those are balls where in your system they are more or less tube segments with chamfers (A and B).

Let’s first look at the springs. Both springs E and F are so-called disc columns consisting of disc springs. That is done to amplify the relative small movement of one single disc spring under compression OR in order to induce a very strong spring force while keeping the movement small. Note that E is the same as F1. The additional spring column F2 makes that each disc spring in spring F has a smaller movement and thus a smaller force than each disc spring in spring E when equally compressed. Therefore spring F is weaker than E. Once the spring rate of a single disc spring is known, we can calculate the spring rates E and F.
Since both springs are captured in a fixed space in mandrel G this mandrel will take a position relative to the middle ring C whereby the spring forces are in balance. The force to move the mandrel G out of this balanced position is distance times the difference between springs rates E and F. The force to move does not differ if the mandrel is moved to the left or to the right.

If we look at the latches A and B. I guess that the dimple in the internal circumference in B is to make sure that each latch has it’s one specific groove. So latch A can freely move over the groove B during firing. And also to make sure that when the mandrel systems reach another set of grooves for A and B somewhere down- or upstream (not visible in this drawing) then it will snap into the correct position again with the latches in the corresponding grooves.

For force calculations I think should not take the dimple chamfers K into account, so latches A and B do not differ in that aspect. Further it looks like middle ring C is wider than the grooves for A and B, so it can pass freely and will not snap in the grooves.

Finally if the mandrel is moved to the left (direction of arrow) the forces on latches A decrease relative more than the forces increase on latches B, so A will snap out of it’s groove rapidly followed in sequence by C and B.

 

[huangcunbing]

Thanks a lot jlnsol!

Your comments and analysis on the latch system give me another perspective and something I missed before.

You know when I keep looking at the same problem and there is not so many close references and resources to help, discussing with others could be the way to break it through.

Here is my confusion you can have some comments on it:

1)when I apply a force, say on either end of the springs, E or F on your marked picture, the lathes are supposed to move into the groove. Ideally,they should mate the groove well. At this point, the two side springs have the same force applied on the latch ring D. Therefore there is no force on bevel contact area between lathes and the grooves. However, it is not the case because if the latches mate the groove well, the preload force for the uplatches (A) is a little bit different from the preload force for the downlatches (B). This difference causes the force unbalance which applied on each side of D. So, there should a force on the bevel contact area to prevent the latches from moving out the mating position. In addition, because the angle is less than the locking angle, even a small forces applied can cause the movement of the latches. In this case, the it will cause the latch to reposition, moving out of the groove a little bit until the spring force is rebalanced. What do you think of ?

2)if the latches mate the groove well by ajusting the springs stifness. That is to achieve the condition that to have the preload force is same for the two springs when the lathes exactly move to its mating position. Then, I apply a force on H, either left or right in attempt to release the lathes. Theoretically and ideally, the latches start to move along the bevel area once the force is applied. During the process, the latches should be further deflected or not. At which point it will be completely released? I am a little bit confused about this at this period.

Thanks again!

 

[jlnsol]

The whole system depends on force transfer by wedges.

Regarding your item 1: The statement “Therefore there is no force on bevel contact area between lathes and the grooves.” is not correct.
To understand this let’s look at the latch ring D which is between latches B and spring F. To balance out this ring D (system is in rest) it is necessary that the force applied on the spring side of latch ring D creates a reaction force on the bevel side of ring D. This reaction force can be resolved in two components: a force perpendicular on the bevel surface and a friction force parallel to the bevel surface. The perpendicular force is the biggest one of the two and is in turn delivered by the beveled corner of the latches B. (For the purpose of understanding we skip the friction for now)

If you then look at latches B. To keep these in balance (system is in rest) the force applied on the right beveled side by ring D should generally be the same as the force on the left beveled side. The last one is delivered by middle ring C. So the latches B undergo mainly two forces on the beveled sides (which can be added as vectors) and try to push them tightly into the groove of tube H. So for sure there is a force on the bevel contact area there.

For proper calculation we should also look at the friction forces. But in general the above can clear up your confusion a bit I think.

In other words: the springs are squeezing the latches out between the middle ring C and the latch rings D into the grooves in tube H. A strong and secured connection is the result.

Regarding your item 2: The statement: “Theoretically and ideally, the latches start to move along the bevel area once the force is applied.” is not correct. What first will happen is the spring E will release and spring F will compress a bit or the other way around depending the direction you move tube H. Look at the last sentence in my previous post about the latches snapping out. At which point the latches will release can only be determined if we take also the friction forces into account. Therefore we should know the friction coefficients and do a full calculation.

Hope this helps.

 

[huangcunbing]

Thanks jlnsol,
I will discuss with you about this later. Busy with another task today.