Episode 074 | Air in Oil with David Placzek and Dr Lukas Hafner (Evamo)

0:00Good day everyone. Welcome to

0:00Lubrication Experts. And today, we are

0:02talking about air in oil. Now,

0:05when I teach classes on lubrication and

0:09lubrication management, we often will

0:11talk about contaminants and the

0:12importance of of contamination as it

0:14relates to oils.

0:17Everyone always thinks of particles when

0:19they think of contamination cuz they're

0:20abrasive. Everyone thinks of water

0:23because it's damaging, it's corrosive,

0:24and it collapses oil films.

0:27Everyone might think of soot, glycol.

0:30Um what is probably less appreciated is

0:33air.

0:35Right, so air as it's entrained in oil,

0:38air as it forms foam, you know, foam is

0:40probably the most visible form of air.

0:43And so people because air is all around

0:45us, people don't think of it necessarily

0:47as being a contaminant. And maybe it

0:49hasn't really occurred to you what the

0:51effects on the oil are.

0:52So, today we're going to dispel all of

0:55those notions. We're going to discuss

0:57and you're going to know more about air

0:59in oil at the end of today than you ever

1:01thought you you ever would. So,

1:04now

1:05as with this podcast, the thing I love

1:08about doing these podcasts is that even

1:11though people think that my job is very

1:12niche as an industrial lubrication

1:14specialist,

1:15there are infinitely more niches within

1:17industrial lubrication. So, amazingly,

1:21yes, there are experts in air in oil.

1:25And I found them, right?

1:27So, this is the best bit of the about

1:28the podcast that I get to drag these

1:29guys and

1:32and you know, bring them up to the

1:33surface to use an air analogy. And and

1:36and we get to highlight the fantastic

1:38work that they're doing in this field.

1:41So, today I've got Lucas and David.

1:45They're from Evamo. And they've got some

1:48really cool technology which we're going

1:49to talk about a little bit later in the

1:50podcast. That's got all to do with with

1:53air and oil. Maybe Lucas, would you mind

1:55giving me a bit of an introduction just

1:57as to you know, who you are and and what

1:59you do in the business?

2:01Yeah, sure. So, thanks for having us.

2:04My name is Lucas.

2:06I have a background in mechanical

2:08engineering as well as in bachelor's and

2:11also master's degree. After that, I

2:14started my PhD with the topic of

2:16production and detection of oil air

2:18dispersions which is basically just the

2:21topic that we are talking about today.

2:23After that, I directly started at Evamo

2:26as a technical lead for the smart bubble

2:28system project.

2:30And also with the industrialization

2:34of what we are doing right at the time

2:36and of what we have been doing

2:40during the PhD.

2:41And now I together with David, we are

2:45let's say

2:46sort of a double arrow in the

2:49go-to-market and product development and

2:51also engineering. And we are very happy

2:54talking today with you

2:56about what we are doing very

2:58passionately.

3:00Yeah, awesome. And David?

3:02Yeah, hi hi Rafe. Also from my side,

3:05thanks for having us. And also as you

3:07mentioned um

3:08giving us the

3:10possibility to bring air in oil more in

3:12the spotlight. So, yeah, my name is

3:15David.

3:16I'm yeah, the complementary part of

3:18Lucas. So, he's more leading the tech

3:20side of the smart bubble system and

3:23everything. And I'm more responsible for

3:25bringing this yeah, technical topics

3:28more into real life value for customers.

3:30So, my background is has nothing to do

3:33with mechanical engineering, lubrication

3:35oil, or something like that. So, I'm

3:37more from

3:39yeah, a software and

3:40IoT business side. So, with some

3:42previous stations within the Bosch

3:44group.

3:45We will maybe also speak a little bit

3:46more about how Evamo founded. So, yeah.

3:51So, basically I'm responsible for

3:52everything that starts with marketing up

3:55to go-to-market and also sharing kind of

3:59yeah, similar topics with Lucas on the

4:02product management side.

4:03Fantastic. Fantastic. So, I found the

4:06two exact right people to to talk about

4:08this topic. So, you know, we've got to

4:10start with the basics.

4:12I think everyone is probably aware of

4:14the different forms of air in oil. But

4:17if you wouldn't mind, could you could

4:19you kind of go through the basics so

4:21that we're all on the same page? So,

4:23just like there are different types of

4:24water in oil. So, you've got you know,

4:26emulsified water, dissolved water, free

4:29water, that kind of thing. What's the

4:31state of play when it comes to air? And

4:34I guess you could you could kind of

4:35relate this to other gases, right? It

4:37doesn't just have to be air, but but but

4:39gases inside of oil.

4:41Yeah.

4:42Yeah, maybe we we we put a lot of

4:44thoughts about how to

4:46provide a yeah, real life technical or

4:49more yeah, tangible

4:51examples for this. So,

4:53maybe to start with a well-known

4:54practical example especially from

4:56Germany is yeah, best example would be a

4:58bottle of beer basically. So, if you

5:00look at to a bottle of beer, you

5:02basically have it all. So,

5:04you have the bottle is closed.

5:06>> [laughter]

5:07>> Here's one I prepared earlier.

5:10I wanted to prepare a beer as well, but

5:12I don't know if it's

5:14No, no, we're good. We're good. Here,

5:15I'll crack one open.

5:17I I hope we will do something in in

5:19Germany if we

5:19>> Yeah, yeah, definitely. This one I mean

5:21this is a Hefeweizen. So, so

5:24>> [laughter]

5:25>> Oh, nice. Nice. Yeah, this is in

5:27Australia as well. Cool.

5:28>> Yeah. Yeah,

5:30as I mentioned, best example would be a

5:32bottle of beer because you have it all.

5:34So, starting with the bottle is closed,

5:36you don't see anything that is visible

5:38in your bottle. And as soon as you open

5:41the lid, the pressure drops. And this is

5:44where the air bubbles for example also

5:46get visible. And if yeah, depending on

5:49how how you handle it before, it's

5:52always starting to foam as well. So,

5:54either in the bottle also passing the

5:56head space.

5:57So, basically coming to lubrication oil

6:00again

6:01looking into yeah, studies and

6:03theoretical stuff is that you always

6:05have 8 to approximately 12% of air

6:09dissolved in your

6:12in your air in your oil. So,

6:15under atmospheric pressure and so on.

6:17So, basically chemically bounded. And

6:19depending on temperature and pressure.

6:22But you have different variants. Maybe

6:24yeah, Lucas can also add something on

6:26this. But in practice, people often say

6:29air in oil. But

6:32yeah, it's not just one thing you have

6:33to do to look into the variant

6:36variables. So, because distinction

6:38matters in air in oil especially. And um

6:42it matters because different air states

6:47causing different consequences. And

6:49there are different ways to react on it

6:51and have a response to it. So, yeah,

6:54basically you have to look into

6:56everything.

6:57Yeah, so just as David has said, we have

7:00like four different types of oils of air

7:04inside

7:05oil reservoir. So, for example,

7:08dissolved air which is

7:10the gas that is physically physically

7:12dissolved in the oil and therefore not

7:15not visible. You can calculate the

7:18amount and the percentage by Henry's

7:20law.

7:21Then of course, the entrained air just

7:23when you open the the lid from the beer

7:25bottle, the discrete bubbles dispersed

7:28in the in the flowing flowing oil.

7:32If those bubbles rise to the surface,

7:34they can form the foam

7:37which is the most stable gas-liquid

7:39structure.

7:40You can then see on the surface if you

7:43take a look at

7:44the surface inside your oil reservoir.

7:47And of course, um

7:49the the head space right above the foam,

7:52right above the

7:54the oil surface.

7:56Um

7:57yeah, which can then exchange with the

8:00fluid by by turbulence and all that

8:05Yeah, awesome.

8:06Um this might seem like a very very

8:09obvious question, but uh

8:11where where does it come from? Because

8:13when we talk about something like like

8:15water in oil, we're always trying to

8:18determine where where is that water

8:19coming from? And it could be it's

8:22process fluids. Let's say it's

8:24steam which leaks past the gland seals.

8:26It could be water that is atmospheric

8:29humidity. It could be wash down water.

8:32It could be you know, just simply rain

8:34getting into the drums.

8:36You know, it seems like an obvious

8:37question, but where does the air come

8:39from?

8:40Yeah, this is this is actually

8:43also something that is on the one hand

8:45side pretty obvious, but on the other

8:47hand side not always that obvious

8:50because the air in the oil, it comes and

8:52it goes. And you have different

8:54possibilities where it can come from and

8:56also how you can influence it

8:59if you can even influence it and not.

9:02So, um one side is the already

9:06air content that is in the oil all the

9:08time. So, at atmospheric pressure

9:10pressure, the dissolved

9:13air in the oil that you can of course

9:15not influence.

9:17But on the other hand side, the

9:18mechanical entrained air in your oil and

9:22also if degassing of this at first

9:25dissolved air content is happening. So,

9:29um

9:30for example, if you take the degassing,

9:33this occurs when the dissolved

9:36air

9:37comes out of your oil. And this happens

9:41when pressure drops, when temperature

9:43also shear conditions change.

9:47On the other hand side, the mechanical

9:48entrainment can come from

9:51for example, if your gears splash inside

9:54a inside a gearbox.

9:57The churning effects,

9:58by aggressive flow conditions inside

10:01your hydraulic system.

10:03If you have a sealant leakage at suction

10:05side of pumps,

10:07or also if you have uh return line

10:09splashing

10:10in your hydraulic systems.

10:12>> And

10:14this is also also um if I can add

10:15something to this. So, um if you look

10:17into particles, for example, of if you

10:19look into water,

10:21um if the oil [clears throat] comes to

10:22your facility, for example, um I think

10:25you can

10:26hope or you can be sure that there's not

10:27there are no particles or water, but in

10:30the field, for example, if we look into

10:31air, air is often seen to be appear out

10:34of nowhere. So, if we're um again

10:36speaking about the air in hydraulic oil,

10:38this is is there anyway?

10:41Um you have the gas often already as um

10:45um

10:45um of of your system, it's part of your

10:47system if you're filling up it up. So,

10:50the operating points um ultimately push

10:53it to

10:54um a more problematic regime in the end.

10:56So, that's why where it comes from is

10:59only half the story, but uh when and

11:01under which conditions it becomes

11:03visible is uh yeah, as as is important.

11:07Mhm. Okay.

11:08Interesting. And then, what about what

11:11about the effects of different types of

11:12oils? So, you know, you've got different

11:15oil chemistries, so different base oils.

11:17Um there's also different additives.

11:19Now, I mean, with my limited knowledge,

11:21I would assume that uh

11:24uh different um

11:25let's say, for example, in refrigeration

11:27lubricants, where we're usually trying

11:29to match the base oil to the type of gas

11:32that's being compressed,

11:33um because you don't want interactions

11:35between the two. So, some gases are more

11:37likely to dissolve in in certain base

11:39oils than others.

11:41Uh and then you've also got surface

11:42tension effects, right? Whether that's

11:44from, you know, the detergent

11:45chemistries or what have you. So,

11:47what is the effect of the actual oil

11:50itself um on the type of air or or the

11:54the nature of let's say the the volume

11:55of air or the the type of air

11:58contamination that that you end up

12:00getting?

12:01Mhm.

12:02Yeah, so this is

12:04in my opinion, one of the of the biggest

12:07effects. Um so, the type of oil and also

12:10the the additive packages that are

12:12inside of the oil. And this is, of

12:15course, all the knowledge the the oil

12:17manufacturers have.

12:19And um

12:21all the time they try to fit the the

12:24performance of the lubricants to the

12:26application they are used for.

12:28So, um

12:31at a very high level,

12:32and also for for modern industry oils,

12:37those are formulated to balance several

12:40functions at once.

12:41So, um

12:43each time with the main goal to fit the

12:45application and the used oil in your

12:48application.

12:49So, for example, if you have high load

12:51carrying and wear protection, you need

12:54to have a lot of oxidation resistance

12:57and thermal stability,

12:59viscosity control with regards to the

13:01temperature inside your application.

13:05Um in the end, if you have a lot of

13:06additives inside your oil,

13:09foam control and the air release

13:11behavior, of course, is is a very very

13:14high

13:15um goal to achieve.

13:17And um

13:19those additives,

13:21they work different and also in

13:23different ways with regards to the air

13:26release behavior. So,

13:29um

13:30for example, if you use the

13:32the the extreme pressure or also

13:34friction modifiers chemistry,

13:37um

13:38those are then used for the mechanical

13:40load handling and also for the friction

13:41behavior. If you take the anti-foam and

13:44deaeration chemistry,

13:46those can handle the the foam then

13:48tendency and also the the gas handling.

13:51If you use the viscosity improvers, they

13:54change the the temperature viscosity

13:56behavior, so how quick also the bubbles

13:58can rise to the surface.

14:01Um and

14:03one of the biggest

14:05points that is really to highlight is

14:07that um

14:09the formulation is always a balancing

14:11act and not

14:13one single objective optimization. So,

14:17um

14:18everything you put inside your oil

14:21um is also changing the surface tension.

14:24And this surface tension is then

14:26influencing how the oil is interacting

14:29with the air inside it. So, it

14:32it influences a lot

14:34um what happens with bubble breakups,

14:37with coalescence behavior,

14:39um what is happening with the surface

14:42activity of the oils and also for the

14:44long-term thermal performance of your

14:47application itself.

14:48So, um

14:51in this case, the air handling of the

14:54oil or of the application is not just a

14:56mechanical question. This is basically

14:59inside the formulation of the oils

15:00itself.

15:03Yeah, and uh yeah, like like Lucas

15:04mentioned, uh so it mostly comes from uh

15:07experience when it comes to lubricant uh

15:10formulation, for example, if yeah, they

15:12are adapting, for example, to a specific

15:14machine or application.

15:16But uh we also found out that two

15:18systems can have have uh very similar

15:20had hardware, but still behave

15:22differently when it comes to

15:24um air ingress.

15:25And uh because then the fluid behaves

15:28differently, um like I don't know, for

15:30example, formulated in the lab. So, not

15:32every, for example, ISO VG 46

15:35um behaves the same once air enters, but

15:38um yeah, that means air handling is,

15:40like Lucas said, not a mechanical topic

15:43always, but um lubricant formulation

15:45topic.

15:46Yeah, interesting. Interesting. Yeah,

15:49I'm you know, it's maybe something

15:50that's a a little underappreciated,

15:52because, you know, and we'll get into it

15:54when we start talking about measuring

15:56air in oil, but people just see a a foam

15:58number and an air release number, and

15:59they kind of call it a day, right? So,

16:02um

16:03uh but I'm sure that it's that's uh

16:04we'll get into that in a little bit more

16:06detail in a second. So, so we've now

16:08established obviously there's there's

16:09different types of air that enter the

16:11oil system. Um you know, we classify

16:14them by by the way that they present

16:15themselves inside of the oil. And I

16:17think everyone will have seen various

16:19versions of this.

16:21Um

16:22then I guess it sort of begs the

16:23question, well, I always say to to my

16:26customers that they should always be

16:27asking, "So, what?"

16:29Right? So, so I have I have oil in uh

16:31let's say, for example, I have

16:32oxidation. So, what? Like, what do I

16:35care about that?

16:36Um so, in this instance,

16:38we have air entering our oil system. So,

16:40whether it's entrained, whether it's

16:43foam, whether it's dissolved air,

16:46so I'm going to ask you the question,

16:47"So, what? Like, what do I care about

16:49that? Why why does it why does it matter

16:51to the performance of the oil? What's

16:52the effects

16:54of that?" Mhm.

16:55Yeah, pretty good question. Um it's

16:58always uh also one thing that we we get

17:01to ask a lot. So, um

17:04we always um

17:07differ between the long-term and also

17:09the short-term effects of the air inside

17:12your system. So, um if you take a look

17:14at the short-term,

17:16um always mechanical damage mechanisms,

17:20they come at first. So, uh if you take a

17:22look at cavitation, microdieseling, also

17:25the efficiency impact of the air inside

17:29your oil, those are very high. So, for

17:31example, if you take the cavitation,

17:33um by increasing the compressibility,

17:37which happens if you put air into your

17:39oil,

17:40um you catalyze the start of a

17:42cavitation. So,

17:44um by reducing the suction height, for

17:47example, this happens a lot by putting

17:50higher air contents into your oil.

17:52Um if you take a look at the

17:53microdieseling, high pressure loads

17:56that uh compress the oil-air mixture and

17:59then produce high temperatures that can

18:02also damage your oil, of course.

18:04Um by hydraulic efficiency, if you

18:07increase the air content, also the

18:10volumetric efficiency of your pumps

18:13isn't the same anymore. So, um

18:16by taking a look at

18:18how much air is inside your system, you

18:20can then take a look at what happens in

18:23short-term.

18:24Um and of course, just as you have said

18:27it at the time, the long-term

18:30um problems are, of course, the

18:32oxidation, because um if you take a look

18:34at uh temperature, catalytic uh metals

18:38or particles,

18:40those

18:41um if this comes together with the

18:42contact surface between the air and the

18:45oil, which is the reactive layer,

18:47um of course, this is only directly

18:50measure measurable.

18:52Um the

18:54the

18:55the thing together is pretty logic. So,

18:58um

18:59increasing the air content also

19:01increases the oxygen availability for

19:04the reaction.

19:06And um

19:07by the air-oil interfacial area, which

19:10defines then the reactive layer for the

19:12oxygen, in combination with the heat

19:15residence time, catalytic surfaces, and

19:19also the contamination,

19:21um the oxidative stress is catalyzed a

19:23lot.

19:25Yeah. Also, you get potential

19:27consequences, including viscosity

19:29shifts, additive depletion, acid

19:32formation, deposits, varnish, and and

19:35all that stuff. So, this is always in in

19:38correlation with with air contents.

19:42Yeah, and just maybe to I got to write

19:44that one um for with regards to

19:47oxidation, for example, oil oil aging,

19:49if we see that that change, for example,

19:51so

19:52um if you do not look into the oil

19:54directly, so really optically you do not

19:56see or can derive cannot revise the oil

19:59air contact surface.

20:01And this is

20:02especially the oil air contact surface

20:04is

20:05is

20:07gaining speed and and and leading in the

20:09end to oil oxidation, oil aging, and

20:12then this ultimately has also an effect

20:14of

20:15yeah with regards to efficiency.

20:17So um

20:19we will talk a lot a little bit more

20:21about that topic, how we are doing it

20:23and how we are doing it optically and

20:25the oil air contact surface is one of

20:27the yeah advantages that is coming out

20:30of our methodology we are using for

20:32detecting air in oil.

20:34Yeah, that's that's so interesting. I

20:35mean, I think

20:37cavitation, micro-dieseling,

20:40you know,

20:41air locks,

20:43loss of hydraulic efficiency. I think

20:45everyone kind of just associates that

20:46with hydraulic units, right?

20:49Um

20:49I mean, I have seen cavitation between

20:52gears.

20:53Uh I think a lot of people might have

20:55seen that those videos on the internet.

20:57Um

20:59and and obviously you can see how those

21:02would translate into other applications.

21:04I think oxidation is maybe a less

21:06obvious one. Um

21:08but it it stands to reason when you

21:10think about it because obviously

21:12at a very basic level oxidation requires

21:14oxygen.

21:15In order to add oxygen to the oil

21:17molecules, you need to have a source of

21:19oxygen and most of the time that's the

21:21air. And so

21:23the less

21:24oxygen exposure you have to the oil, the

21:26less oxidation happens.

21:28But maybe it's a it's an

21:29under-appreciated

21:31fact.

21:32>> That that's that's two two good examples

21:34for that. So every everybody knows it

21:37for example from a banana. If you look

21:38into a banana and then

21:40um you're getting it out and it has

21:43contact to to the air, for example, then

21:45it gets yellow or brown and the same

21:48actually happens also to the oil. So it

21:51ages with contact to to the oxygen or

21:54gas or anything else.

21:55And

21:56then coming to my second example,

21:59everybody knows how

22:01hard it is to to swim, for example, but

22:04if you compare now for example swimming

22:07swimming through water, it's something

22:08different like swimming to honey. So

22:12if you have a more contact bigger

22:13contact surface uh

22:17um the oil ages faster

22:19then the viscosity is changing and then

22:21you need more input for getting the same

22:23output. So for example, if we're talking

22:24about wind turbine gearboxes.

22:27Yeah. Yep, interesting. So yeah, so I

22:30think

22:32yeah again, it's one of those things

22:33where maybe this is under-appreciated of

22:36the the effects that

22:38that we can have between

22:40air and the oil system.

22:42So so we've established

22:45what kind of air

22:46actually gets into the oil.

22:48And then now I think we've given enough

22:51evidence to say

22:53can be a bad thing, right? In some cases

22:55it's going to be unavoidable, but but

22:56but can be a bad thing. So now let's

22:59talk about the techniques that we have

23:00in our toolbox

23:02to reduce the amount of exposure that we

23:05have between the air and the oil system.

23:06So so what can we do to prevent foam?

23:10What can we do to prevent air

23:12entrainment? What Ultimately, what we're

23:14trying to prevent is is the is the

23:16damage through oxidation, cavitation,

23:18micro-dieseling, loss of hydraulic

23:20efficiency, etc. So so what are the the

23:22tools that we have in our toolbox in

23:24order to accomplish that?

23:26Yeah, we quite often hear that question.

23:30So the industry is often looking for a

23:32silver bullet, but air in oil is rather

23:35a

23:36one variable problem.

23:38Um so you usually need to look into a

23:41layered approach. So on the one hand

23:42side as we already

23:44um

23:45described it, we have the hardware side.

23:47So there are different techniques for

23:49reducing like more well-known changing

23:52the reservoir design,

23:54optimizing the return line geometry,

23:56de-aeration zones, and maybe also coming

23:59to a practical example from one of our

24:01projects is the shaft suction condition.

24:04So for example, we had a

24:06measurement request for a pump

24:09manufacturer

24:10from Austria and they are providing

24:12their pumps for example for submarines.

24:14So it's obvious that you do not want

24:17noise and you do not want to have any

24:19wear

24:20or damages in in your system.

24:22And so they always tried to optimize

24:26their design of their pumps with regards

24:28to cavitation by really changing the

24:32design and geometry of the pump.

24:34But they never looked deeply into the

24:36suction condition. So um with with our

24:40test benches where we either have the

24:41chance to reproduce the air content, so

24:44we really can ingress it in a controlled

24:46manner, we could then found out that

24:49their specs that they defined for

24:51withstanding so how much air

24:53can be in their pumps was totally wrong.

24:56So they declared or specified their

24:58pumps with 9% for example and we we

25:01approved it that it's not even reaching

25:035% of of air in oil until it starts to

25:06cavitate. So they really followed the

25:10wrong technique since more than 5 to 10

25:12years and yeah but because they didn't

25:15measure it. So every time they had a

25:17issue in the field, they sent out a

25:20service mechanism

25:22for example or technician.

25:23And

25:25they flew to the problem to the

25:26application, changed the pump and so on

25:28and so forth.

25:29So

25:30yeah, the suction condition for example

25:32and the return line geometry is

25:33something really important. So on the

25:36one the other hand side we have the

25:37fluid. So for example, the technique

25:40would be

25:42better understanding and changing the

25:44viscosity for example or we have

25:46influences in the air release behavior

25:48and

25:49also for example additives like

25:51anti-foam chemistry that is used to

25:54reduce the effects of air in oil.

25:56So in the end it's usually really a

25:59combination of the fluid system of the

26:02fluid the chemistry, the system design

26:04and also the operating conditions.

26:07So yeah.

26:09I think that's that's that are the most

26:11well-known but also I don't know tips

26:14from from our side.

26:16Yeah, interesting. Really interesting. I

26:19mean, there's obviously some some some

26:21other like

26:23chemistry tools that we have in in our

26:25toolbox in terms of you know, anti-foam

26:27agents and obviously you can affect air

26:29release properties by

26:32but even even as basic as as the the

26:34viscosity of the oil, right? So high

26:36viscosity oils are going to release air

26:38bubbles a lot slower.

26:40Yeah. Can you can you talk also I mean,

26:42one thing that you just talked about was

26:44the

26:45basically suction line design.

26:48Right? What about the effect of the

26:50actual reservoir itself?

26:53So I mean, the way that I've always

26:54taught it to to students is like you if

26:58you had two choices, right? So every

26:59everything else holding every other

27:01other thing equal, we've got two systems

27:03that are identical and in one you've got

27:06a reservoir that is

27:08you know, same volume but wide and

27:10shallow and you've got another reservoir

27:12which is tall and skinny.

27:14You know, which one do you think is

27:15going to generate more foam?

27:17Just as a bit of a thought exercise.

27:20Now that's obviously taking it to the

27:21extreme.

27:23But what kind of best practices I guess

27:25do you guys see in terms of

27:27reservoir tank design?

27:31Yeah, so just as you as you mentioned,

27:35I think the

27:36the worst case would be a very high

27:38tank,

27:39very high

27:41oil reservoir with uh

27:43suction and pressure inlet on the same

27:46height um with the same um

27:50volume flow. This would be pretty bad.

27:53So

27:54the things that that would work quite

27:57well is if you take just a look at the

28:00the height the bubbles have to

28:02um

28:03have to

28:04walk through to get to the surface to

28:07get out of the system and also the

28:10distance they have to travel to get back

28:13into the suction line. Um if you take

28:16those few variables, it can get pretty

28:19easy for you to optimize such a system.

28:22Also if you take

28:24a look at the amount of um

28:28of oil that is sucked into your system

28:30and also fed back Yeah, interesting.

28:33Interesting. So so

28:36if I had to kind of

28:38if I'm an if I'm an end user, right? So

28:40and I'm looking to troubleshoot, we need

28:43standardized methods to be able to

28:45measure

28:46air in oil.

28:48Um to my knowledge at the moment like

28:51when

28:52when I look at different test methods,

28:54I'm basically looking at foam sequence

28:56testing and air release testing.

28:58Right? And and that's basically the

29:01extent of the toolkit. So could you

29:03please help us understand you know, what

29:06those are for?

29:07And now obviously you guys have

29:09developed

29:11a system outside of that. So what was

29:14the

29:15what was the gap that you saw or what

29:17what what about the current tests like

29:20makes it incomplete?

29:23Mhm.

29:24Yeah, so this is always the thing

29:27as you have mentioned the air release

29:29behavior of the oil so also the foam

29:31tendency with the three sequences.

29:33Um this is always the the point of view

29:36from from lab.

29:37So what we are also doing is taking a

29:40look at the field. So what is really

29:42happening with the oil inside the

29:44application and

29:45um on that point we have different types

29:48of measurement possibilities.

29:51Um, on one hand side, you have the

29:54indirect measurement

29:56methods. So, if you take a look at, for

29:59example, electrical constants like the

30:02electricity, capacitive, also

30:04impedance-based methods.

30:07Uh, you can track a change change of

30:10your of your fluid or of the mixture

30:14inside your measurement device and refer

30:17this to an air content in your inside

30:19your system.

30:21Um, you can also use light-based

30:22methods, acoustic response-based

30:25indicators, and all that stuff.

30:28Um, this was also the first way to go

30:31um, in our

30:33um, development, but uh, we could really

30:37not find a way to reproduce the

30:39measures, but also to to really validate

30:42optically what was happening inside our

30:45systems.

30:46So, this was the point to really start

30:49with a direct measurement method.

30:51Um, and this is what we are doing. So,

30:54we are are really taking pictures of the

30:56oil-air mixture and then evaluating

31:00um, each and every bubble that that we

31:03can measure. Yeah, interesting. Uh, can

31:06I just uh, uh, step back cuz you

31:08mentioned impedance type sensors or kind

31:11of you know, electrical property

31:12sensors. This is something that I've

31:14seen a fair bit, um, you know, not

31:16adjust for identifying air content in

31:19oil, but also oil degradation,

31:22uh, contamination, all that sort of

31:24stuff. I mean, um,

31:26uh, there's a there's a few of these

31:27sensors, whether it's impedance or

31:29resistivity. Uh, and it's always struck

31:31me that you're getting a single number

31:33out, right? It's, you know, kind of

31:34electrical property up or down. Yeah.

31:37But, but

31:38like water should affect that and air

31:40should affect it and contaminants should

31:42affect it and oil degradation could

31:45affect it. You you know, in in the

31:47conception of how these sensors work,

31:49how are you identifying a single failure

31:51mode? Or or can it not be done?

31:55This is this is the exact problem we

31:57also see here, um, because, um, how do

32:00you want to interpret a value that is

32:02changing by so many uh, side effects?

32:06So, um,

32:07um, when we started the project, we also

32:10started with with those um, electrical

32:12constant measurement devices

32:14and

32:16we had more to do with the calibration

32:18of those sensors

32:20than with the measurement itself. So,

32:23this was

32:24this was very crazy and um, therefore,

32:27it was it was really not

32:30for us a thing to

32:32to develop a new type of sensor system

32:35to integrate into into applications, but

32:38really to solve a an issue that we had.

32:42Yeah, and the issue is not only is not

32:43only with regards to data or to

32:45technology, but

32:47yeah, with regards to my role of

32:48functional function is

32:50yeah, the more the gap with regards to

32:52the customer value. So, the field

32:54problem is is highly dynamic, but many

32:57methods like Lucas already mentioned,

33:00um, are indirect, for example, or

33:02difficult to interpret physically under

33:04changing conditions, for example. So,

33:06for us, it was really important to not

33:08just provide um, a test bench that can

33:11use can be used in one environment, but

33:14from lab to testing to field and and

33:16even to simulation. So, um, and this is

33:19quite hard with the current um, yeah,

33:22technology that that are on market

33:23because the issue is not that the

33:25current methods are useless.

33:27And um, the issue is more that they are

33:30often not close to not closing the loop

33:32between symptoms, the actual air states

33:35in the field especially, and also the

33:37engineering action. So, currently, if

33:39you look into, for example,

33:40die-dielectricity

33:42technologies, uh, you just get one um,

33:45variable or KPI like you mentioned, but

33:47you don't do not know what to do with

33:50it. So, for example, how to derive some

33:52engineering actions.

33:54So, the market does not lack sensors and

33:56data, but

33:57we are calling it really one a one data

34:00language. So, we can use our system in

34:03the lab. We can for early days testing,

34:05for example, and we can use it in the

34:07testing. If you, for example, want to

34:08specify a specific lubricant for your

34:11pump,

34:13and we can also then um, get some

34:15results from the field how the pump is

34:17actually working um, or behaving with

34:19this specific lubricant.

34:21So, um, this I think is the gap that we

34:24are closing especially not from also not

34:26only from a technical point of view, but

34:28really when it comes to customer value

34:30in the end.

34:32Yeah, that's that's really interesting.

34:33I mean,

34:35like you said, you know, closing the gap

34:36between between the lab and and and what

34:38what the reality of the situation is.

34:40The example that I always kind of give

34:42to to clients and customers and and

34:45students is you you know, as an example

34:48would be foam,

34:49right? Um, so, people their hair seems

34:53to set on fire for some reason when it

34:55comes to foam,

34:56uh, even though it's probably a little

34:58hopefully be more benign than entrained

34:59air, but but uh,

35:03the statement that I always kind of give

35:04them,

35:05maybe you guys can tell me if I'm

35:06telling them the wrong thing, but I

35:07always say, you don't have

35:10a foam problem until you have a problem

35:13with foam,

35:14right? And that seems like a an odd

35:16thing to say to them, but but what I

35:19mean by that is

35:20um, just because the lab result is

35:24telling you that foam sequence is

35:26degraded, right? Compared to the new

35:28oil, that doesn't necessarily mean that

35:30you actually have a foam problem. Like,

35:32go and look at the reservoir.

35:34You know, if there's no foam, you're

35:36fine.

35:37Uh, just because the test tells you that

35:39there is the potential for foam, uh,

35:42doesn't necessarily mean that you run

35:43into a problem in the real world. And

35:45even if you have foam, that's like the

35:47oil naturally releasing the air. So,

35:50until you have problems, you know,

35:52controlling the level of the oil in the

35:54sight glass or you know, the foam is

35:56getting sucked into the the suction side

35:58of the pump or the the foam is spilling

36:00out um,

36:02of the of the reservoir,

36:05you know, you don't actually have a

36:06problem yet. Um, and so, it's good to

36:09see it's good to see someone who's who's

36:11looking to close the loop. So, speaking

36:13of, right? You guys have developed a new

36:15system

36:17um, for I guess measurement and and

36:19modeling of air in oil. You hinted

36:22already that it's kind of an image-based

36:24system and we're able to to to look at

36:27different air bubbles and categorize

36:29them, but could you please just go into

36:31a little bit more detail on like how it

36:32works and and maybe where did it come

36:34from, too? Cuz, you know, all

36:36technologies kind of have a an origin

36:38story for want of a better word. Yeah.

36:41Mhm.

36:41Yeah, so, um, just as we we've said

36:44before, the the origin of the of the

36:47measurement device itself, um, was more

36:50practical than a problem-related

36:53topic. So, um, within the pump business,

36:57um, from Iwam we had

36:59a lot of

37:01um, tests we had to do with the pumps

37:04related to the air inside transmission

37:07systems.

37:09So, customers always said, you have to

37:11guarantee that your pumps withstand like

37:1540 to 50% of air, sometimes 30% of air,

37:18which is pretty crazy if you really take

37:21a look at gearboxes.

37:23Um, and

37:25there we tested several

37:27um, devices for measuring, also several

37:30possibilities for for reproducing and

37:34none of them really worked for us

37:36because of um,

37:38cross-contamination,

37:40particles, water content, all that stuff

37:43that you have already mentioned before,

37:45that you already know from from your uh,

37:47side of view.

37:49Um, and because to repro-

37:51reproducibility,

37:53um,

37:54was really not sufficient and also we

37:58had a a constant need of calibration of

38:01our sensors.

38:02Um,

38:03we had to

38:06we really had to um, find a way for us

38:09to develop such a such a measurement

38:12system.

38:13And um,

38:14this is also or this was the point where

38:17I I started my PhD and so, um, the

38:21original problem was not

38:24that we really needed a new sensor.

38:26Yeah.

38:27Yeah, but yeah, maybe I I can add

38:29something with regards to Iwam. So, um,

38:32yeah, basically, we we both work on the

38:34smart bubble system, but the history of

38:37Iwam is quite interesting also with

38:38regards to this topic because um, for 5

38:41years yeah, so,

38:43yeah, ex-

38:44yeah, 10 10 10 days before the the fifth

38:46birthday. So, 5 years ago, uh, we got

38:49bought um, from an investor from Unic

38:52and we're a former Bosch business unit,

38:55um, yeah, producing, manufacturing, and

38:57selling um, steering pumps and

38:59transmission pumps. So, um, in the CV

39:01sector, so commercial vehicle steering

39:03um, pumps, we are market-leading

39:05company, um,

39:07especially in the western hemisphere.

39:09And um, but Iwam itself is is quite

39:11unknown, uh, but to be honest, they're a

39:14hidden champion in in that way. So, um,

39:17as Lucas mentioned, um, it really came

39:19out of an own problem. So, we didn't

39:21develop a solution out of nowhere, but

39:23we really had the same issue

39:26um, like the others that we can see now.

39:28So, the problem was not, as mentioned,

39:30um, to buy another sensor that we can

39:32test, but the real problem was really

39:33the incomplete visibility of uh, what is

39:36happening in our pump. Um, and that we

39:39needed for decision-making. So, the team

39:41had lab test like you mentioned um, we

39:44had field we had field testing with had

39:48a sensor for the testing but there was

39:50still the gap between we suspect air and

39:52we really can quantify the air bubbles

39:54in our system. And that matters really

39:57commercially also with regards to to

40:00hidden cost factors to hidden fluid side

40:03variables that can inside or can can sit

40:06inside a very expensive downtime chain.

40:09So

40:11this is this is also also with regards

40:13to what I mentioned to the different

40:15environments is

40:16that

40:18you also mentioned it yourself so we do

40:20not want to generate a next sensor zoos

40:23because there are so many sensors on the

40:24market but the gap is really that if

40:27you're for example using the 9120

40:30so the

40:31air release ISO test in the lab so

40:34you're measuring it for example with

40:35density but in the testing you're then

40:38using a dielectricity sensor and in the

40:41field you do it something like different

40:44so we want to really

40:46establish and bring the more spotlight

40:48into air and oil but not only for the

40:51field where the volume is but also to

40:53really

40:54to really show that we are understanding

40:56the air and oil phenomena in different

40:57environments so really starting with lab

41:00going to the testing and also to the

41:04field.

41:05Yeah so maybe explain how the system

41:07works right because you've already

41:08mentioned that it's like image based

41:10right so we're we're we're capturing

41:12actual images of the air bubbles but I

41:14think maybe

41:16it might be a little bit difficult for

41:18for some viewers to understand what it

41:20means to be able to take the same

41:21technology which is image based and and

41:24what does it mean to have it in the lab

41:26but also in the field like what

41:29what is the technology and how are you

41:31looking to deploy it?

41:33Yeah so at a very high level

41:36we're basically sucking oil out of the

41:40the oil reservoir and then we define

41:43specially defined measurement path

41:45controlled optics and specially designed

41:47illumination concepts

41:49yeah to make the dispersed air

41:55structures visible

41:57and this with high

41:59consistency and accuracy and

42:04just like this you can really

42:07have a reproducible but also

42:09understandable result so that you don't

42:12have to interpret and think about yeah

42:15maybe my my water content got also a

42:17little up and so my the electricity

42:20changed a little bit we have to

42:22recalibrate and all that stuff so we

42:24just plug it to the application suck the

42:27oil out measure it and

42:30that's that's mainly it.

42:32And this is really the advantage because

42:34you mentioned that

42:35currently with the current technologies

42:37you just have one single data so it's

42:41just for example if we would looking

42:42into the air release behavior with the

42:44lab environment again you have the any

42:46density topic and you have the time

42:50which which is measured but we are not

42:52only looking into air content or air

42:55volume because you can either have 5% of

42:57air content in your oil with with

43:00smaller bubbles and but you can also

43:02have 5% of air volume in your oil with

43:05smaller bubbles.

43:06So this really depends and this is also

43:08the the advantage of our system in the

43:10end because we are not only detecting or

43:12counting the bubbles

43:14but we are looking really into every

43:16each and every air bubble in in your

43:17system or in your oil so we are counting

43:20them we are classifying them or modeling

43:23them and yeah we have the size

43:25distribution we have the contact surface

43:29of each and every air bubble and into

43:31the fluid

43:32and so this really makes the difference

43:35if we look into

43:38into measuring the air in in the field

43:40and then we can really use that one data

43:42language

43:44to have more clear insights into what is

43:47happening in your lubricant but also in

43:49the hardware system

43:51and

43:52the key difference here just to add this

43:54one as well is that we're not just

43:56measuring a proxy that tells you

43:58something has changed in the system or

43:59something gets noisy or something like

44:01that we really try to make the the the

44:06topic of air or air bubbles

44:08more observable but in the end and this

44:11is the most important topic is more

44:13explainable and actionable so instead of

44:16just saying the the system sounds wrong

44:18or

44:20I don't know the vibration is

44:22influenced or impacted you can really

44:24say see and say how the air is behaving

44:27in different systems but all

44:35detect but also interpret

44:37and actionability is one of the key

44:39points here.

44:41Yeah interesting so so

44:43you know when whenever I'm explaining

44:45condition monitoring I always say uh

44:47you know a a blood test is only useful

44:49if you act

44:51on

44:52the data right so if you if you get a

44:54blood test and it says you have high

44:55cholesterol and you continue to not

44:58exercise and and eat the same rubbish

45:01food that you always have then

45:04nothing's changed right

45:06so so that's the key thing right you you

45:07mentioned the word actionable.

45:09So so what are the kind of things that

45:12customers would typically be doing in

45:14response to the data that they get from

45:17the smart bubble system so

45:19you know you mentioned it obliquely

45:20before wind turbines so I've got a wind

45:23turbine I I I take these measurements of

45:26the air bubbles you're giving me a lot

45:28of data on you know contact surface

45:32you're giving me the size of bubbles the

45:34distribution of the size of the bubbles

45:36and the total volume as well.

45:39Okay now what do I do with that those

45:41numbers?

45:44Yeah this is is a very

45:47interesting thing because

45:49each time you want to optimize something

45:52you have to know what you really want to

45:53optimize so you have different paths

45:55that you can go so for example if you

45:58want to do a thermal

46:00optimization

46:01if you put air in inside of your oil the

46:06thermal conductivity is changing a lot

46:08so if you have a lot of isolating air

46:11inside your oil the cooling needs a lot

46:13of energy.

46:15When we are then at the energy if you

46:17put a lot of air inside your oil the

46:21viscous effects regarding to to power

46:25loss is also changing a lot because on

46:27the one hand side your filling level

46:29changes therefore the resistance of your

46:31gears inside the gearbox is changing a

46:33lot

46:34on the other hand side if you want to

46:36change the long term

46:38issues

46:40with regards to to air content for

46:42example the oxidation the temperature

46:45and also the contact surface between the

46:47air and the oil is influencing this a

46:49lot.

46:50Also how your oil is performing with

46:53regards to

46:55temperature and air contents so we have

46:58plenty different of measurements

47:01where you can really see

47:03the behavior of the oil

47:06with regards to the air release behavior

47:10with temperatures so not each oil is

47:14performing the same so with with

47:16increasing the temperature the air is

47:17getting out much more quicker some oils

47:20with some additives can then keep the

47:23air much better because the the air

47:26bubbles get much smaller.

47:28So those are the topics that you really

47:30have to define with the customer

47:33and then you can point at them very

47:36easily and then you can also optimize

47:38them the way that you want to do it

47:41but only with a measurement and this is

47:43this is the key point.

47:46Yeah interesting so so I mean the types

47:47of optimizations that you're talking

47:49about I guess probably weren't possible

47:53in the past right because we've never

47:54had

47:55that kind of data

47:57to be able to play with

47:59and I suspect you know it's it's one of

48:02those situations where

48:03when you measure the data you find uses

48:06for it as well

48:08maybe that were a little unexpected so

48:12So so one good example is for example if

48:14you are for example looking to the air

48:15release value measurement

48:17methodology or techniques that they are

48:19currently established

48:20they are missing quite most of the

48:24the most important

48:26sections of the diagram so if you look

48:28for example into the ISO 90 9120 you

48:31just get the line that is decreasing and

48:34then it tells you how much it how much

48:36time it took to get to the 0.2

48:39percentage of air density

48:41and

48:42we are really really looking not into

48:44this declining

48:46line but also the beginning the peak and

48:50the first 10 10 10 seconds for example

48:53so this is really giving you then the

48:55advantage here so it's more not only

48:58more trustworthy but more honest

48:59technology so because you can really see

49:02what is happening and if you're not

49:03trusting your diagram for example you

49:05can see which which kind of image which

49:08kind of condition were there or air

49:10regime when you

49:12when you

49:13measured your lubricant for example.

49:16So

49:17and if you look for example in the field

49:19we can also provide a dynamic air

49:21release value so then you can correlate

49:23what is happening in the lab in the

49:25field but also you can then correlate

49:27the air release behavior with the oil

49:30oil air contact surface

49:32so

49:33rather you have a air release that is

49:35happening fast

49:37seeing also the air contact surface or

49:39air contact surface that is then showing

49:41you do you have more like bigger bubbles

49:43that are the cause of the good air

49:45release or do you have still smaller

49:47bubbles that are more

49:49harmful to your system and also

49:51infecting or influencing the the air

49:53release behavior.

49:55And and how you seeing the systems

49:58deployed for the most part, right? So I

50:00can imagine the way that you've

50:03described it is that you're taking oil

50:05from the reservoir

50:07obviously in a very controlled manner

50:09in order to be able to do the

50:10measurement. So in a sense, that's kind

50:13of an inline unit, right? Because it's

50:15giving you effectively real-time data.

50:18Is that

50:20usually going to be you know attached

50:22just for diagnostic purposes and then

50:25that gives us a window into the current

50:27state and then we can go and make fixes

50:29or are you guys looking at this as a

50:32more permanent solution? Like how how

50:33are they being deployed?

50:36Yeah, this this is this is a good

50:37question interesting topic. So

50:40maybe strange that coming from a more

50:42business-related person but we do not

50:45want to scale it just for more devices

50:47for example.

50:48So we really want to see that or show

50:51that we are really

50:53putting the spotlight on air in oil

50:56getting to know it more better

50:58providing more insights into it

51:01really also working on the

51:02standardization topics side for

51:05example and so and so and so forth. So

51:07it was really important for us not to

51:09just develop a

51:11another sensor for example that is maybe

51:13on the market already and

51:15nobody is trusting sensors anymore and

51:17so on. So we really wanted to make sure

51:19that we before we enter the market and

51:22make more scalability in terms of

51:24devices more devices. We really want to

51:26understand the topic and not only for

51:28the field where the volume is later on

51:30but for the lab for the testing for the

51:32field and also for the simulation. So

51:35for us scale means not only more devices

51:37but a robust method that is also

51:41reproducible and and stable in terms of

51:43quality

51:44and also consistent in the evaluation

51:47logic. So once again if you're speaking

51:50about different languages in terms of

51:53what is the measurement technology

51:56what is the output in terms of data, you

51:58really need to see that the logic of the

52:00evaluation of the data that you're

52:02gathering

52:03is the same across different devices in

52:06your portfolio.

52:07So um for us it's really turning the

52:11technically strong method into something

52:13that fits real engineering workflows for

52:15example. And

52:18yeah maybe also talking about different

52:20markets and use cases. We also speak

52:23about with simulation companies for

52:26example because they also some gaps with

52:29missing real data in the gearbox. So

52:32everybody knows the glass front or glass

52:34covered gearboxes that are used as a

52:36foundation for simulation in the end but

52:40yeah you can you can easy easy it up or

52:42simplify this process by just putting

52:45real data into it. So this is why we

52:47really speaking from lab to field and

52:50simulation and not only in lubricants or

52:52oil. So last week we were there

52:54analytica or two weeks ago in Munich and

52:58we're even speaking about biotech use

53:00cases where air bubbles are

53:03are needed for example to to create some

53:06plants. So

53:08we really want to see what the effect of

53:11the air is not only in lubricants or oil

53:13but in different in different markets as

53:16well because either you want it or you

53:18want to don't want it. So if we are

53:21getting back to the example of the beer,

53:22you obviously want to have the foaming

53:25contrary to to the lubricant market. So

53:28the broader mission is really to move

53:30the topic of air from an invisible issue

53:33to something that is yeah more

53:34measurable, more discussable and also

53:37more extendable.

53:39Awesome. Well,

53:41um

53:42I think that's a great place to end it.

53:44You know I promised at the at the very

53:46at the very top of this

53:48this podcast that the viewers would

53:50leave with a better understanding and

53:52appreciation for air in oil and I

53:55certainly feel like I am leaving with a

53:57better appreciation for the subject. So

54:00hopefully the viewers are too because

54:02there's a hell of a lot more to it um

54:05than than meets the eye. So to both of

54:07you David and Lucas, really appreciate

54:10you coming to to discuss you know your

54:12your particular niche and the great

54:15contributions that you're making as

54:16well.

54:17I'm really excited actually to see the

54:19unit potentially in person a little bit

54:21later in the year and

54:25really appreciate your time.

54:27Thank you. Thank you for having having

54:29us. Maybe also to add this so where you

54:31can see it and how we can use it or to

54:32deploying it at the moment. So we are

54:34offering pilot measurements because you

54:36have obviously it's a new topic or not a

54:38new topic or but a different approach

54:41and we want to provide the possibility

54:43to just prove the value for each and

54:44every application. So

54:46we are offering the pilot measurements

54:48on one hand side and maybe you probably

54:51saw it on LinkedIn. We have a very cool

54:54demonstrator that we are using in our

54:56exhibitions and conferences. So the next

54:59one will be dry test where we also

55:00showcase the combination of real

55:03measurement data with simulation. So you

55:06can stay

55:07tuned on this as well and on the

55:09lubricant expo in Dusseldorf where you

55:10will be also as an exhibitor and

55:13surprisingly we have

55:15the the booth contrary to the to your

55:17booth. So hopefully you will have a lot

55:19of guys that want to visit you in

55:21Germany. Yeah, that that sounds that

55:24sounds good. I'll make sure to put

55:25details for you guys in the description

55:28below and so if anyone wants to to reach

55:31out you can reach out that way.

55:34All right, thanks gents. Really

55:35appreciate it.