Page 1 on 26

Propulsion of a boat by means of a pop-pop engine

Contribution to the knowledge of the pop-pop engine.

Warning: This document was written in its original version in Februar 2!!". #ome complements and corrections have been made but it has not been completel updated. #ee the documents written later.

$ast updating % Feb.12& 2!1!

Page 2 on 26

Summary !. Forewords 1. 'rchitecture of a pop-pop engine. 2. (escription. ). *here does the "pop-pop" sound come from+ ,. #mall reminders of mechanics and thermod namics ,.1. Thermal engine ,.2. Ph sical characteristics of the water ,.). -inetic energ ,.,. .verheated steam ". /ow does the pop-pop engine propel the boat+ 6. #implified version of the running of the pop-pop engine 0. Factors influencing performance 1. 'nal sis of what e2ists 1.1 3easurements 1.2. 42periments 1.). (eductions 1.).1. Fre5uenc . 1.).2. $i5uid piston stroke. 1.).). 4fficienc 1.).,. Pipe temperature 1.).". Pressures. 6. $aws of ph sics7mathematical models 6.1. (rum 6.2. Pipe 6.). 8o99le 6.,. :oat 1!. Probable response ;due to lack of measuring tools< 11. 'dditional measurements and results. 12. (etailed description of the running of the pop-pop engine 1). *h does the boat vibrate+ 1).1. /elmholt9 resonator+ 1).2. #imple resonator. 1).). Two mass resonator. 1).,. Change over from one to the other one. 1).". Can we use the energ of a resonator+ 1,. *hat kind of flow in the pipe+ 1". The ideal pop-pop engine+ 'nne2 1. #ound generator characteristics.

$ast updating % Feb.12& 2!1!

Page ) on 26 0. Forewords *ith no special interest at that time = had seen pop-pop engines for decades. =n >anuar 2!!" one of m children brought back to me from =ndia the to whose picture is given on the front page. = reall en?o ed this to @ from the scientific point of view. Curiosit & availabilit ;?ust retired<& adapted scholarship and industrial e2perience about thermal engines do the rest. $ooking at tens of *eb sites ;in French and 4nglish< related to pop-pop engines shows that most of them have mostl a mercantile content or purpose. .ne can see mainl small to boats built abroad& notabl in =ndia and sold in 4urope 1! to "! times more than locall . ' few sites of enthusiastic amateurs are more interesting. Finall & a limited number of sites show a scientific aspect. Anfortunatel the given e2planations ;/elmolt9 resonator& #tirling engine& Bankine or other Carnot c cle@< have onl a remote connection with our sub?ect& or are incomplete. 't least these sites do e2ist& and ?oining forces@ Though the 5uantification of the phenomenon of ph sics that makes the pop-pop engine works is difficult& its description is nevertheless eas ;cf. C6<. /owever none of the e2amined sites e2plain wh at a given time of the c cle a vacuum is created in the boiler. Two basic scientific notions are generall missing% kinetic energ and overheated steam. =n addition& man other details are not mentioned. *e are going to tr to fill this lack with e2planations understandable b an one. (espite the fact that the efficienc of a pop-pop engine is ver ver bad& knowing the problems of water 5ualit and water treatment in a water-steam c cle it is unrealistic to foresee building a big pop-pop engine for an industrial application because the water that it uses is the one on which the boat is sailing. 8evertheless the topic D even for to s manufacturing D is interesting. Therefore& for those who are willing to go deep in it we will suppl a calculation method and the formulas that appl to relevant phenomena of ph sics and thermod namics. There is much to doE #ome recommended sites% www.eclecticspace.net www.scienceto maker.org7boat7inde2.htm ;the best< www.che9.com7llegoff7poppop ;has a picture of the same =ndian boat as mine< www.galepp.com7boat7popboat.htm ;idem< http%77ourworld.compuserve.com7homepage7?pFperroud7poppop.htm http%77membres.l cos.fr7moudge ' last site interesting for the histor and an eccentric point of view& but which is disgusting for its mercantile aspect% www.pop-pop.fr . :ut& whatever& if =8P= ;French office which registers the brand names< accepted in 6" to register Gbateau H moteur pop-popI ;pop-pop engine boat< for the e2clusive usage of a privileged compan & it is mostl =8P= that is to be blamed because this group of words was known and used for decades. =t is as if toda a car seller would register Gcar with a diesel engineI as a brand name to tr to dominate the market. This present document has no commercial issue& nevertheless& to avoid an misunderstanding this said registered group of words will not be reused.

$ast updating % Feb.12& 2!1!

Page , on 26 1. Architecture of a pop-pop engine.

(rum

#tern of the boat

*ater level

:urner

Pipe

:ottom of the boat . !escription. Though there are man kinds& the pop-pop engine is a model of simplicit . There is not an moving part. The most common ones use two parallel pipes to ease the filling& but to understand how it works& onl one suffices. ;This will be demonstrated.< This pipe which sends water towards the boat stern ;pulsated water?et effect< is supplied b a steam drum. This one is a drum in name onl . =t doesnJt have the usual shape of a boiler steam drum because at the same time it is used as a drum and as a heating place& and must have a large internal surface to ease the steaming. 'nd& in order to generate a pop-pop sound the drum includes a deformable metallic diaphragm. The heat needed for steaming comes generall from a candle or a ver small pan filled with alcohol. 'ccording to one of the e2amined *eb sites there is a model improved b using of a condenser. The condenser is a simple absorbent tissue damped with cold water and set on the pipe near the drum. 't the same time the tissue is supposed to be a bilge pump thanks to capillarit . #ome people dare to sa thatE ". Where does the "pop-pop" sound come from# $etJs now sa a few words about this in order not to come back on this matter. The well known sound is due to the deformation of a small and thin metallic membrane ;as the top of some preserve cans<& sometimes conve2& sometimes concave. This metallic membrane is an integral part of the drum and is distorted b the c clic pressure variations inside the drum. The sound level increases with the suddenness of the change between concave and conve2. To improve this& at rest the membrane could be concave with a slight strain. The sound is something that can be heard b spectators& but it is not needed for the good working of the engine. .n the contrar E The variation of the steam drum volume due to the deformation of the membrane decreases the efficienc of the engine. :ut the efficienc of the pop-pop engine seems not to be a concern for an bod ;e2cept on one siteK<. For additional info about the sound generator of the to we have studied see anne2 1.
8oteK% .nl one of the e2amined *eb sites sa s something about efficienc . 'nd what an efficienc E 6!LE 6!L of what+ -nowing the works of so man smart scientists for decades to reach an efficienc of about "!L on a surpercharged diesel engine& and far less on a turbine one must be dreaming. #ee C1.).).

$ast updating % Feb.12& 2!1!

Page " on 26 $. Small reminders of mechanics and thermodynamics: $.1. %hermal engine. Can onl work between two heat sources having different temperatures. There are man theoretical ;simplified< c cles% Carnot& Bankine& :eau de Bochas& #abathM& #tirling& $enoir& (iesel@ 8one of them corresponds to our application. :ut we alwa s find , main steps% 1. Compression of a gas 2. 'ddition of heat ). 42pansion and production of mechanical energ ,. Cooling $. . Physical characteristics of the water: 't standard atmospheric pressure and at 1!!NC& the specific densit of water is 6"1kg7m)& the steam one is !&"6kg7m). Therefore& a drop of water changed into steam takes 16"! times its initial volume. $.". &inetic energy: 'n moving ob?ect having a speed O ;in m7s< and a mass m ;in kg< is characteri9ed b 1 2 a kinetic energ 4 ;in >oules< which is defined b E = mV . 2 'n speed change of this mobile re5uires a transfer of energ between it and the outside. The term object is to be understood here in a broad sense. =t is not necessaril a solid. For our application it will be a li5uid% the water inside the pipe. $.$. Saturated or o'erheated steam: 4ver bod knows the improperl called saturated steam. =t is the one we see ?ust above the saucepan or at the pressure cooker output when the water boils. =n fact it is visible because it is a mi2ture of steam ;gas< and micro-droplets of water ;li5uid<. =t e2ists another kind of steam which is not so known& nor even easil guessed% it is the overheated or superheated steam. =t is not visible. =t is commonl used to suppl steam turbinesP for instance at 6! bars and "1"NC& though the boiling temperature at that pressure is onl ;E< 20"NC. *hen a mass of steam is overheated its energ ;it is called enthalp < is increased& which doesnJt mean that the pressure is changed. For instance& at standard atmospheric pressure and 1!!NC the enthalp of the saturated steam is 2602 k>7kg& but if at the same pressure the temperature is increased to 1"!NC& the volume is multiplied b 1.1) ;ratio of the temperatures in N-< and the enthalp becomes 2000 k>7kg. *hat must be understood to follow is that it is possible simultaneousl to increase the temperature and decrease the overheated steam pressure. That could be obtained b heating a container of variable volume& for instance a c linder provided with a piston. (uring the heating process& the temperature and the enthalp can increase though the pressure can decrease because the piston is moved. =n a pop-pop engine there is no metallic piston& but the free surface of the water inside the pipe moves and acts as a piston.

$ast updating % Feb.12& 2!1!

Page 6 on 26

#team vessel

$i5uid piston

.verheated steam #aturated steam

*ater

(. )ow does the pop-pop engine propel the boat# The water located inside the pipe is alternatel pushed and pulled ;sucked< b the drum. *e will see later wh . From that point& ) as mmetrical phenomena contribute or could contribute to propel the boat. 1N< The hull of the boat has an as mmetrical profile which facilitates its forward move when submitted to alternate solicitations. Certainl this is not the most important factor. *hen we rock between ahead and astern on a dingh with sharp bow and stern transom we succeed in making it move forward& but using the same energ to handle a scull oar gives a 5uite better effect. 2N< The alternate movement of the water inside the pipe is not s mmetrical. =ndeed& water can move faster towards the stern than forward because the effective depression inside the drum cannot e2ceed the steam limit ;!&!2) bar' at sea level with standard atmospheric pressure<. There& once more& one can doubt the effect of this as mmetr e2cept in one case described in C11. )N< The water flow at the e2tremit of the pipe is not reversible. =n the propulsion phase the water is pushed astern ;propulsion b pulsed water?et<. =n the rela2ing phase the water comes from an direction. This is something which is well described on man *eb sites. P*+P,-S.+/
Pipe *ater?et

*0-A1A%.+/
Pipe

$ast updating % Feb.12& 2!1!

Page 0 on 26

' comparison can be done with a two stroke engine which is characteri9ed b a period of propulsion during the e2pansion of the gasses& and a period of braking D not so much D during the compression phase. To eliminate an a priori a test has been performed on our =ndian to b fitting two elbows at the outlet of the pipes in order to aim the ?ets sidewa s. Besult% phenomena 1 and 2 have no effect. The boat was no longer sailing in spite of a good working of the engine. For those who would be interested& we have written a small specific document to e2plain the Gworking principle of a pulsed water?etI. 2. *unning 3simplified4 of the pop-pop engine: The chronological break down is as follows% =nitial situation% drum and pipe full of water. Firing of the burner. Oapori9ation of the water inside the drum. The steam pushes the water into the pipe. The water snake located in the pipe is in progress. The drum contains onl a remaining of steam that is being overheated. The water snake D due to its inertia D continues to move and creates a partial vacuum in the drum. The vacuum is improved b the fact the steam moving in the pipe cools down and condenses. This slows the li5uid piston& and then reverses the movement. The water reaches the drum and is transformed 5uasi-instantaneousl into steam when touching the metal. @. 'nd this& until the flames e2tinguishes. Then the drum cools down& the steam condenses and we come back to the initial situation with drum and pipe full of water. This 5ualitative description illustrates perfectl how the pop-pop engine works but nothing is 5uantified& and if it is eas to ?ustif the permanent conditions it is more difficult to e2plain wh or how it starts. /ow can we succeed in reaching during the first seconds a speed of water sufficient enough to create a vacuum in the drum+ The answer is not obvious. =ndeed& as soon as boiling begins steam pushes water ;at appro2imatel 1!!NC<. 's the steam gets further in the pipe it cools down& on one hand b contact with the pipe& and on the other hand b conduction and mi2ing with the one met further down in the pipe. This cooling involves condensing of the steam located close to the separation surface. This pulls up the water@ *e will tr to give a more detailed e2planation in chapter 12. .ur =ndian pop-pop engine seems to find its cruise fre5uenc after between 2 and ) seconds& but the amplitude of the vibrations increases during the 2 or ) following seconds. 'bout the amplitude of the water movements inside the pipe& intuitivel one can think that the e2haust of the steam is to be avoided& but the best result should correspond to the renewing of practicall all the water at each c cle. Therefore& there is a good compromise to find between the heating power and the si9es of this propulsion plant.

$ast updating % Feb.12& 2!1!

Page 1 on 26 5. Factors influencing the performances: Temperature and power of the hot source Temperature ;and power< of the cold source 'reas and thermal e2change coefficients #hape of the drum (iaphragm softness Position and shape of the drum-pipe ?unction Thermal inertia of the drum Thermal inertia of the pipe $ength of the pipe (iameter of the pipe (iameter and profile of the no99le ;the orifice< =t can be added that for the adaptation to the propulsion of a boat some other factors pla a role% /ull profile (isplacement ;mass< Position of the center of gravit =nclination of the pipe /ow deep in the water is the no99le /eight of the drum from the water Theoreticall we also should take into account the h drod namic pressure due to the boat speed& but this one is ver small and conse5uentl negligible. 6. Analysis of what e7ists: *hat was learned form the tests D up to now non destructive D performed on a to % 6.1 8easurements: - :oat mass% )!g ;including the engine& but empt < - (rum mass% about ,g - Thickness of the heated part% !&,mm ;including soft metal< - 3aterial% steel ;tin-plate< e2cept for the membrane which is made of brass. - =nternal diameter of the 2 pipes% ).)mm - $ength of each pipe% 16mm - Total volume ;drum Q pipes<% about 2.2cm) 6. . 07periments: - The fre5uenc of the c cle is 5uasi independent of the heating powerP above a certain minimum. #ome tests were performed with power ratio appro2 from 1 to 1!. - The delivered mechanical power increases with the heating power. This is visible on vibrations and amplitude of the generated waves. =t is clearl felt when the boat is held in place b a hand. - The to speed ;from !.2m7s ahead to !.2m7s asternP forced b hand< has practicall no influence on the fre5uenc . - The geometrical height of the drum from the water level ;between 2 and 6cm b lifting the bow of the boat< doesnJt influence significantl the fre5uenc and doesnJt disturb the pop-pop generation. - Plugging one of the pipes involves a decreasing of the fre5uenc . 's soon as it is unplugged it works as previousl . - $engthening of the pipes b "!L doesnJt change significantl the fre5uenc .

$ast updating % Feb.12& 2!1!

Page 6 on 26 - 4ven after a long period of running the pipes are cold where the fingers can touch them. - The GcruiseI fre5uenc is 0 to 1 /9. - 't ver low power ;when the candle goes out but while the wick is still red< the engine is still generating a pop-pop but ou must be ver close to it to ear it because the membrane is no longer moving as before. - The heating power that we used ;F< was about 21*. ;The heat source is a small birthda candle& the power of which was evaluated b heating a well known water 5uantit and measuring the temperature increase versus time. Pictures of the test and file of the measures are available.< - The Gcruising speedI O is appro2 !.1"m7s ;!.",km7h< - Towing the boat at that speed re5uires a pulling force of appro2 2m8 ;2 milli8ewtons<. =t is minute. #o minute that it is the measure on which the uncertaint was the worst one. *e performed man additional tests with better measuring tools to improve itJs knowledge. 6.". !eductions: 1.).1. Fre5uenc . =t seems not influenced b most of the parameters. =t is something as for a pendulum the amplitude of which can be changed easil & but the fre5uenc of which is constant so long as neither the moving mass nor rope length is changed. 1.).2. $i5uid piston stroke. :ecause of the propulsion principle& the boat being sailing at the speed O& the output of the water?et needs to be at least once par c cle faster than O.. vma2RO $etJs suppose the flow perfectl sinusoidal. 4ver point of the li5uid snake moves b d = aSIN t = aSIN 2 Ft F being the fre5uenc of the c cle. (erivation of this e5uation gives the speed of the water inside the pipe. v = 2 FaCOS 2 Ft vma2S2FaRO. *ith FS1/9 and OS!.1"m7s we get aR)&2.1!-)m aR)&2mm a being the half amplitude of the displacement& the total stroke is more than 6.,mm. 1.).). 4fficienc % These last three data allow to compute the global efficienc which is the ratio between the released mechanical power ;drag force multiplied b speed< and the supplied heating TV power which is the one of the candle% r = For this application F !.!!2 !.1" r= = !.!!11L . =tJs patheticE 's the measures were not performed as laborator 21 ones& the relative uncertaint on some of them is bigP but the are onl three. 'ssuming we were ver bad or ver unluck so that on the three of them we made an error in the same wa & and from single to double ;it is nevertheless enormous to do that<& the efficienc would become !.!!11L. *e can accept that during the GenduranceI test the flame of the candle was not as big as it could have been& and if we accept a worse condition 1! times less& the best result could onl reach !.!11L. =t is to be compared to the )"L of a classic propulsion ;"!L for the engine and 0!L for the propeller<. This is still ver ver bad and ?ustifies that there has been no industrial application of the pop-pop engine.

$ast updating % Feb.12& 2!1!

Page 1! on 26 Comparison of a propulsion b pop-pop engine with a mechanical propulsion ;springQpropeller< on to s of the same si9e. 1N< #pring. To wind up the spring re5uires about 1! turns of the ke with a tor5ue of 2!!m8m ;!.28m<. Corresponding energ % !.2K2KpK1!S12.6>oules. =t is minute. This engine propels the boat for appro2 12.6 seconds ;to simplif <. Therefore& the power is 1*. Taking into account the mechanical efficienc and the one of the propeller the delivered power is smaller. $etJs sa !."*. 2N< Candle. Though it is not obvious& the power and the energ delivered b a candle are relativel big. ' small birthda candle ;mass% 1 gram< delivers appro2 )"* as heat& and it takes 1! minutes to burn. Corresponding energ % 21k>. *ith 2 grams burnt in " minutes ;data from Professor $e :ot< it means ,2k> and 1,!*. )N< 4fficienc . The efficienc of such a small to is likel about 1! times less than the one of a big shipP i.e. appro2 )."L. Professor $e :ot measured similar thrusts with mechanical propulsion ;1* ).1!-28< and pop-pop propulsion ;1,!* 1.1.1!-28<. =t means a ratio of 2)) in favor of the mechanical propulsion. (ividing )."L b 2)) gives !.!1"L and it can be checked that it is lower than !.!11L calculated before b e2cess. This consolidates our measures and computations. 8ote% specific e2periments ran in 2!!6 with an electric heating source allowed to improve the knowledge of the efficienc and to improve the efficienc in some circumstances. /owever& it remains ver bad. #ee the document entitled G4fficienc of a pop-pop engineI. 1.).,. Pipe temperature% The efficienc being what it is& nearl all ;more than 66L< is heatP that is to sa increase of the water temperature and of the surrounding air. 't this step some data are missing. $etJs assume that the amplitude of the water oscillation in the pipes is ,!mm ;the onl certitude is that 6.,TTaT16<. The volume of water renewed at each c cle is D2 V = ,! 2 = !.61cm) . 't 0 to 1 /9 this corresponds to an average flow of " cm )7s. , This water comes from the surroundings in which the boat sails& for instance at 2!NC it F 21 = = 1.) C and becomes 21.)NC. =n fact& due to the increases b = Q c " 1! ) ,11" partial renewing of the water& and due to permanent agitation& the temperature is ver likel progressivel higher when approaching the drum& and this is in accordance with the fact the pipes seem cold when we touch them where the are accessible. This is based on a plausible but not verified h pothesis. =f we use a pessimistic one corresponding to a stroke of 1!mm instead of ,! ;ratio ,<& the temperature becomes 2".)NC. =tJs still cold for the fingers. *e will come back on that further in this report. 1.).". Pressures. =n operation& the effective pressure in the drum fluctuates at 1/9 between a minimum lower than -2!mm*U and a ma2imum higher than 66Q,!S1!6mm*U.
8ote% 3ini and ma2i pressures were measured with some other engines. The show that the absolute value of the low pressure is alwa s lower than the one of the high pressure.

$ast updating % Feb.12& 2!1!

Page 11 on 26 1.).6. 3assic power. *hen in operation the total weight of the engine ;drum Q pipes Q water< is appro2imatel 1 grams ;e2cluding fuel<. This engine delivers a mechanical power of !.)m* ;2m82!.1"m7s<. /ence& the mass power is )0."*7ton. $etJs compare with a pleasure boat. For )0."k* ;"1/P< the total weight of engineQreduction gearQshaftQpropeller is appro2imatel 2"!kg. Therefore& the mass power of our pop-pop engine is about "!! times weaker than the one of a classic propulsion plant. =t is pathetic. To be honest we must sa that industrial life learned us that e2trapolations bring sometimes ama9ing results. The scaling factor between the to and the pleasure boat is so big that the truth could be better& but the probabilit to e5ual classic propulsion is ver low. 'nd we let ou imagine how much fuel would have to be carried. 9. -aws of physics:8athematical models: Bemarks% 1N< The best efficienc of a propulsion b water?et corresponds to an outlet water speed which is ver close to the double of the one of the boatP this water being thrown in the air ;hori9ontall <. :ut this principle cannot appl here because the pipe must imperativel be in the water in order not to suck air.
8ote 1% =n the first release of this document we wrote what follows in italic% To convince you that it is less efficient you could ma e a little e!"eriment# Set yourself in the $arden %ith the %aterin$ hose and a buc et of %ater# To hold the hose no&&le you need to e!ert a certain force in the direction %ere the %ater $oes# '(ou counter the "ro"ulsive force of a %aterjet#) No%* "ut it inside the %ater of the buc et* and you %ill notice that the effort to maintain the hose no&&le is far less# This is sub?ective and wrong. =n the air the ?et noise is louder but the thrust is roughl the same. *e did a specific e2periment to measure this. 8ote 2% 3athematicall one demonstrate that the best efficienc of an aircraft ?et corresponds to an outlet speed of the gasses which is ver close to the one of the aircraft. For a boat the problem differs because the water which is sucked is not stead . The comparative demonstration is available upon re5uest.

2N< The pop-pop engine is disconcertingl simple& but to calculate such a motor re5uires computing tools which did not e2ist when it was invented. *e face a periodical phenomenon in biphasic environment ;water and steam< which is ver comple2. To anal 9e it ou could simpl ;soft euphemism< write the e5uations at time GtI to determine the values of the parameters at time GtQtI& and do it again for GtQ2 tI@ =tJs a stupid ?ob that a PC is able to do ver well and fast insofar as all the algorithms and initial conditions are given to it. )N< Thanks to relativel high fre5uenc of the pulsations generated b the pop-pop engine ;generall several hert9& sometimes several tens of hert9< some parameters can be considered as constantP which ease a little bit the problem. *e will note G/nI the simplif ing h potheses which seem sensible to use. =t remains to write the e5uations that appl to an subs stem& and to make them interact. *e will not go up to the resolution of the problem because there is an infinite number of possibilities depending on the materials used& the geometr of the parts& the power of the heat source@ .ur purpose is ?ust to show that a modeling is possible& and hence it could be possible to optimi9e a pop-pop engine& for instance b adapting the heat power and7or the length of the pipes to the water temperature.

$ast updating % Feb.12& 2!1!

Page 12 on 26 6.1. (rum =t receives a calorific flow ;a power< that can be considered as constant. 'lso b convection and radiation it loses a power that can be considered as constant. The resulting power F is used to heat the drum and the water in it or to overheat the remaining steam. ;For connoisseurs& look at 3ollierJs diagram.< /1% the heat power F is constant. The c cle anal sis will be 5uite longer than the real c cleE To simplif & we will split it in three phases% - Filling of the drum. *ater goes from pipe to drum. - *ater e2haust from the drum. =t goes to the pipe. - #team overheating. There is no li5uid in the drum. 6.1.1. Filling of the drum. The heating power is transmitted to the drum and to what is inside.
d t d F = Qt .c; t T1 < + + + mt c t + dmt dt dt

=n this formula& the first term corresponds to the temperature increase of the water entering the drum& the second corresponds to the temperature variation of the metal of the drum& the third to the temperature rise of the water inside the drum& the fourth and last term to the steaming of part of the water. The notations are the following ones% Vt S flow as volume of water entering the drum at time t. S specific gravit of the water. ;=t varies a little with the temperature& but we will use S 1!!! kg7m)< c S heat coefficient for water. ;=t varies onl a ver few with the temperature. *e will use cS ,11" >7kgNC< t S temperature of the water inside the drum at time t. T1 S temperature of the water entering the drum. ;=f the engine is si9ed properl practicall all the water is renewed at each c cle. =t can be considered that T1 is the temperature of the water on which the boat is sailing. For instance% 1"NC.< 3 S 3etallic mass of the drum S 3ass heat of the metal. ;=t varies a little with the temperature& but we will use ,6! >7kgNC for steel and ,!! >7kgNC for copper.< mt S mass of water as li5uid in the drum at time t Pt S pressure in the drum at time t. ;To limit the risks of errors& use ever where in the calculations the absolute pressure.< S vapori9ation heat of water at temperature t. =n addition& during this drum filling phase& the content is biphasic and pressure and temperature are linked b the saturation curve ;3ollierJs diagram<. This allows to calculate P;tQt< and ;tQt< from Pt. and t. Therefore& we can track the evolution of the pressure in the drum at each computing step. 6.1.2. *ater e2haust from the drum.

$ast updating % Feb.12& 2!1!

Page 1) on 26 's the drum contains simultaneousl water and steam& it will be assumed that this latter is saturated. The formula is similar to the previous one& but the term corresponding to the water that was entering no longer e2ists. F = +
d t dt + mt c d t dt + dmt .

=n addition& there is an e5uation for the mass variation. (uring the time interval t some of the water goes from the drum to the pipe. :ut this mSmt-m;tQt< will onl be known at the occasion of the whole computation including interacts between drum and pipe. (uring the phase of water e2haust from the drum& we can calculate P;tQt< and ;tQt< from Pt. and t. 6.1.). #team overheating. =n this phase there is no more li5uid water in the drum. (ue to the fact the pop-pop engine works at relativel low pressures the steam mass ;taking into account its specific heat coefficient< is minute and negligible compared to the one of the drum itself. /2. =n this phase the steam mass can be neglected. Therefore& the formula becomes F = +
d t dt

. The temperature rise is linear. The thermal

energ stored during this phase will be restored later when the water comes back 6.2. Pipe =t contains a 5uantit of water and a 5uantit of steam that are changing. (ue to a relativel good heat transfer coefficient of water and bad steam one& it can be checked e2perimentall that the pipe remains cold when the engine is GshakingI. 's a confirmation& some pop-pop engines were built& using plastic straws as pipes. Conse5uentl & it can be assumed that there is no heat transfer through the pipe above where it is in contact with the surrounding waterP but there is condensation of some steam b contact with the water film which is created along the inner wall of the pipe when the water goes down. The vacuum being created on one hand b the condensation and on the other hand b the displacement of water upon inertia effect& one must find a good compromise between the 1 2 kinetic energ ; E = mV < and the friction losses ;BSkO2< that is to sa between diameter 2 and length of the pipe. =t is obvious that an hitch ;sudden bend& diameter change& ringed fle2ible hose@< is to be avoided. For a given kinetic energ & the friction losses are minimi9ed b % - Ase of onl one pipe instead of 2 or more. - Circular section of the pipe - Pipe without a hitch between drum and no99le - Uood surface roughness inside the pipe.
8ote% =t is ama9ing to see that some of these basic precautions were not taken into account on man pop-pop boats.

The specific gravit and the viscosit of the water are 5uite different from those of the steam. /)% The friction losses of the steam inside the pipe can be neglected compared with the ones of the water. =nside the pipe there is a water snake of mass m that is submitted upstream to the absolute pressure of the drum& and downstream to the atmospheric pressure increased b the

$ast updating % Feb.12& 2!1!

Page 1, on 26 draught of the no99le& and b the pressure drop because this snake is moving and there are friction losses against the pipe internal wall. =n addition to all that we have to take into account the fact the pipe is more or less inclined. 8otations% - # S area of the pipe cross section - $tS length of the water snake at time t - mt S its mass at time t - h S geometrical height between no99le and water surface - S inclination angle of the pipe - Ot S velocit of the snake center of gravit - t S acceleration of this center of gravit - Vt S flow as volume through a cross section of the pipe PS drum pressure

$72

mg

P!SPaQgh

OtS172 Vt7# ;P-P!<#Qmt.g.sinSmtt m;tQt<S mt-Vt.t ;$et chose a convention for the sign and stick to itE< 6.). 8o99le =t is the more or less narrow end piece fitted to the pipe. =ts purpose is to increase the speed of the water coming out. #1
O1

#2

O2

,1 V 12 , 2 V 2 2 + = + 2$ 2$ 'nd& the li5uid being considered incompressible we get% V 1 S1 = V 2 S 2 The application of :ernoulli theorem allows to write% #hould the pulsation be not so fre5uent& one could imagine an improvement consisting to put in parallel a no99le and a bigger orifice provided with a check valve ;non return valve<&
$ast updating % Feb.12& 2!1!

Page 1" on 26 and even to add also a check valve on the no99le in order to push water astern and to suck on the fore. This deserves to be tested for e2ample with check valves of the Gduck beakI t pe which can work at rather high fre5uencies. #ee the following scheme. Thus& the rela2ation phase would be propulsive as well and the efficienc could be 5uite improved. The pop-pop engine would be the prime mover of a reciprocating pump ;without piston e2cept a li5uid one< used to propel the boat.
8ote% The drawings are not as good as we would have e2pected& but b easiness and to give ever bod an eas access the have not been designed with sophisticated C'( software.

P o p p o p

#ucking

42haust

8ote% For those who would be interested to do such an e2periment& for the manufacturing a piece of thin plastic hose with a soft ironing on one end can do a cheap check valve with a ver small inertia. $etJs come back to what e2istsE =n practice it is ver seldom that there is a no99le because a reduction of the cross section means an increase of the friction losses and conse5uentl % - ' decrease of the kinetic energ created b the vacuum - 'n increase of the rela2ation timeP hence of the c cle period. /,% For the first calculations& it will be considered that it is the end of the pipe itself which is the no99le. ;8o cross section change<. 6.,. :oat =t sails at a speed that can be considered as constant because of the ratio between its relativel big mass ;e2% several tens of grams< and the mean value of the mass of water in movement in the pop-pop engine ;e2% an average of 1&, grams<. *hen the boat is stopped ;bollard pull test< there is no noticeable change of the c cle fre5uenc . /"% The speed of the boat is constant and too slow to influence the performances of the poppop engine.

$ast updating % Feb.12& 2!1!

Page 16 on 26 To validate all these simplified h potheses& an e2isting pop-pop engine is to be modeli9ed and the final results of the computations are to be compared with the real measures. 10. Probable response 3due to lac; of measuring tools4 'ccording to all what has been seen or heard& the response of a pop-pop engine should be similar to what follows.
ins tantane ous e ffective pres sure in the drum 140 120 100 80 Pressure in mmWG 60 40 20 0 0 -20 -40 -60 Tim e in s e conds
Tim e delay (heating to 1 00 C)

fir t !o!

tart of reg"lar !o!-!o! r#g"lier

't tS! the fire is lit 't tS2 the drum reaches 1!!NC 't tS2.6 the first pop can be heard. Then there is one pop per c cle 't tS)., the pop-pop becomes regular ;2 pops per c cle<. 't tS" the engine delivers practicall its full power. Times and pressures are given for indication onl . The can var with man parameters. 11. Additional measurements and results. Five pop-pop engines ;2 small ones and ) bigger ones< each one using a single pipe have been first built and tested with different materials& several diameters& and several lengths for this pipe. Pictures and file of the measures are available. 't each new test onl one parameter was changed. 4ngines nN1 and 2 had man similar characteristics ;mass& volume& heating surface<. 4ngines nN2& ) and , had big drums ;2,.6cc< as the goal was to get low fre5uencies for better looking at the hunting. Then other engines have been built. The knowledge from all this is what follows% =t is confirmed that one pipe onl suffices. ;/owever it is more difficult to fill.< The height of the drum from the water level is not a ma?or criterion. =t changes slightl the boiling pressure and temperature inside the drum& but there is no visible change of the rest ;within the limits of the pipe length<. #uccessful tests were done in 2!!" with the drum up to )! centimeters above the water level surface& and down to 1"
$ast updating % Feb.12& 2!1!

Page 10 on 26 centimeters. #ince that time we e2ceeded 2 meters above and 1! meters below ;thanks to a pressuri9ed tank<. ' good thermal conduction between the drum and the pipe degrades the performances. This is due on one hand to the fact that the overheating is more difficult and on the other hand to the fact that the boiling is not so sudden. To interpose a short isolating connection improves things. The drum must be located more or less at the top of the engine& and an inclination of the pipe is needed in the area where is located the interface. The length of the pipe has little influence on the fre5uenc of the hunting. *ith a short pipe the hunting is difficult to get and not regular& sometimes impossible to get. The fre5uenc increases with the heating power& but not much. The amplitude of the reciprocating movement of the water in the pipe increases with the heating power. For a given volume& starting and maintaining the hunting is easier and the amplitudes are bigger when a drum with a large vapori9ation area is used. The fre5uenc seems related onl to the volume of the drum and the diameter of the pipe in its part where there is steam in it. The formula could be something as n F= -o$ ;1." K 2 + 1 in which O represents the volume of the drum in cc and V the diameter in mm& and n the number of pipes. 'ccording to our first estimates& the coefficient could be ".). 'dditional tests with laborator means would be necessar to define more accuratel or to find a more relevant formula. =n the coil engines ;with the pipe itself coiled in the middle and heated at that place< it is difficult to sa where the drum ends and where the pipe begins to appl the formula. *hat is for sure is that the volume which is used as drum is small and that the resulting fre5uenc is high. 3ost of these fabrications are made of copperP which has the inconvenient of a good thermal conduction between the hot part and the cold one. .ne of the fabrications ;cf. www.eclecticspace.net < was made of glass. For what concerns the thermal conduction it is rather bad ;S1*7mN- while for copper it is S)6!*7mN-<& but to look at what happens it is e2cellent. =t is noted a ver long time to start& and then a high fre5uenc ;2"/9< due to a small volume of what works as a drum. *hen the pipe is made of thermall isolating material ;plastic< and is narrow& the water temperature increases regularl from the no99le to the drum. *hen a no99le is fitted where the pipe is connected to the drum& the length and the diameter of the pipe have no influence insofar as the pipe diameter is larger than the one of the no99le. The fre5uenc is then onl a function of the drum and of the length and diameter of the no99le. The alternate movement of the separation surface between water and steam e2pressed as volume ;diameter of the pipe multiplied b the stroke of the li5uid piston< seems onl linked to the heating power. 'ir inside the drum increases the performances ;fre5uenc and amplitude<. 't cruising speed the free surface of the water inside the pipe is more or less distant from the drum. The water never enters the drum on its li5uid state. =t is the interface between saturated steam ;more or less titrated< and overheated steam which goes in and out of the drum. There are onl micro droplets. =f there was one drop ;!.!"cc< of water& when vapori9ing it would occup appro2 12cc and would through out all the water from the drum and from the pipe. *hen the heating power is removed the engine continues to run ;for 1" to )! seconds for our first two protot pes< due to the thermal inertia of the drum. The separation

$ast updating % Feb.12& 2!1!

Page 11 on 26 surface between water and steam comes closer to the drum. =f some water is poured on the drum it boils. =t is a proof that the temperature is more than 1!!NC. 'nd& when the temperature comes down to appro2 1!!NC it can be seen and heard that the water climbs up suddenl through the pipe into the drum. =t is a proof that the drum was filled with overheated steam. =f the drum is naturall cooled down& it takes a longer time& but the result is the same. *hen one drop of water enters into the drum it is because the temperature is less than 1!!NC and the whole steam collapse. There is a water hammer effect which propels the boat. ' minimum heating power is re5uired to start the hunting. :elow this& there is a static e5uilibrium with the free surface of the water ver close to the drum. For the same heating power we noticed a ratio of about 12 between the momentum of the first engine ;diameter 6& stroke 2"& fre5uenc ,< and the one of the second engine ;diameter )& stroke 1!& fre5uenc ).)<. Conse5uentl & the surface of the pipe cross section is to be chosen as large as possible@ but it seems that there are practical limits for the ration #7O ;section of the pipe7volume of gasP this volume being bigger than the drum one when the interface is located inside the pipe<. These limits are 1 S 1 < < appro2imatel such that to build an engine which can work we need 2!!! V 1!! with # in mm2 and O in mm). #ome tests were performed with the addition of a condenser. 8o improvement was seen. *hen an engine works well ;rather high fre5uenc and long stroke<& the water in the pipe is so shaken that its temperature is progressivel higher when approaching the drum& but still cold enough not to vapori9e inside the pipe. ' series of tests at bollard pull conditions has been carried out to determine the thrust developed b several pop-pop engines& an of them supplied b a small candle. .ur =ndian to ;2 pipes of diameter ).)< is in the top four. *e have got similar or slightl better thrust with engines using onl one pipe of% o (iameter 6 with a no99le of diameter ".2. ;Transient test. 8ot confirmed<. o (iameter ".2 with a no99le of diameter ,.,. o (iameter ,.2 without no99le. #everal web sites refer to a burning out phenomenon which could be due to some air ingress. (onJt confuse air and overheated steamE #ome of our tests lasted several tens of minutes without encountering this problem. :ut& if ou wait for a time long enough& ou can observe that in addition to the well known /F hunting ;, /9 for our first engine& ).) for the second< there is sometimes a $F one ;period 2" seconds for our first engine& 22! seconds for the second one< which is far from s mmetric. =f in these conditions the pipe is big there is periodicall a violent water?et& and then the water comes back slowl to the drum. =f the pipe is not big there is a steam blast& and then the water comes back slowl to the drum& and suddenl fills it ;when it is cooled b the first drops of water<. Uraphicall the result ;pressure in the drum versus time< looks like what follows. The dead 9ones correspond to the time needed to increase the temperature of the water which filled the drum from cold to 1!!NC.

$ast updating % Feb.12& 2!1!

Page 16 on 26

150,00

100,00

50,00

0,00

-50,00

-100,00

-150,00

Concerning the presence of air& some GaccidentalI tests or ver long ones ;several hours< were enlightening. 1N< *ith some air the fre5uenc and the amplitude are higher. 2N< 'ir is removed automaticall in case of burning out. *h + *e have no certitudes& but strong presumptions based on the fact that we saw ver clearl & on some occasions when the engine refused to start hunting ;shaking<& that the gas ;air and steam< was pushing down the whole water column and bubbles were finall escaping ;at 1! to 1"cm below the water surface< and going up to the surface. =f it had been steam it would have collapse. 'fter having though it through& air has an atomic mass of appro2 26 though the steam one is appro2 11. #o& air is heavier than steam and accumulates at the bottom when there is no hunting. =t could be the same when the engine is hunting. *hatever it is& the presence of air ;accumulated or mi2ed with steam< modifies the heat transfer between steam and water film. 1 . !etailed description of the running of the pop-pop engine. To look at the phenomenon ) different t pes of GtransparentI engines have been built. Those of >eff ;who believes that ever thing is to be seen in the drum< use a drum with a transparent plastic top cover. $oWc made one with a glass coil to see ever thing. The engines of >ean-Xves ;who believes that ever thing is to be seen inside the pipe< use a transparent plastic pipe. 4ach of those who had pre?udices ;antagonist< has got confirmation that he was right. *h + :ecause the engines could not be compared. The performance depends on the materials& on the heating power& and on the geometr of the engine. >effJs engines use a drum with a thin lower wall which is not ver good heat transferring& and an isolating top cover& and pipes made of copper. The heat source is located at the place where the flame touches the drum. The cold source climbs up to the top of the pipes. Besult% when >eff looks at the drum he sees ever thing in it. The steam bubble rises ?ust above the flame and its si9e varies at the same fre5uenc as the pop-pop. >ean-XvesJ engines have a massive drum made of copper and a pipe made of isolating material. The whole drum is submitted to the heat source while the cold source is far down in the pipe. Besult% the ;oscillating< steam7water interface is located inside the pipe. (uring some of our tests with small diameter pipe the position of the interface was measured to be down to 2!cm from the drum. $oWcJs engine is isolated ever where. The interface is created somewhere between the heated place and the cold source. 'nd the hunting e2ists around this mean position.
$ast updating % Feb.12& 2!1!

Page 2! on 26

#ince that time& >oao Cordero& Christophe& $oWc& Uuus and >ean-Xves built engines made of glass& and ever e2periment completes or confirms the other ones. =n spite of the visible differences& in an of these engines the process is nevertheless the same. 's soon as boiling starts& steam pushes water. 't one end of the steam pocket there is overheating and volume increase& while at the other end the steam is cooled down b contact with the surrounding material and with the free surface of the water. This cooling down involves steam desuperheating with volume reduction& and condensation of the steam located close to the water surface. 'nd there& depending on the circumstances and on the engines& one can observe ver different phenomena versus time. /unting could be either immediate& or it could occur after a long GhesitationI dela ;sometimes more than )! seconds<. This latter case is the result of a 5uasi e5uilibrium between the volume increase created b the overheating of the steam located close to the heat source& and the volume decrease involved b the desuperheating and the condensation on the cold side& because in this slow D even ver slow D process& the kinetic energ is negligible. *hen the steam7water interface reverses its wa & it goes 5uickl towards the heat source. There the inertia pla s a non negligible role. The volume of the steam pocket decreases& while the heat source increases its energ . The pressure has got two reasons to increase. This pressure increase slows the water column progress& and reverses it. The steam that goes awa from the heat source cools down@ The reciprocating movement is fed until the heat source disappears. /ote 1% b feeding >effJs engine with a power bigger than usual we succeeded to make it work as ours with a Gdr I drum. Though it didnJt last long ;because the plastic cover melted< this confirms if it had been needed that the same laws of mechanics and thermod namics appl . The mean volume of the steam pocket depends ;among others< on the heating power. /ote % on our engines the temperature of the thick copper drum is usuall between 11! and 1)!NC. The steam7water interface never goes into the drum. The heat is e2changed b conduction& convection and radiation between various steam la ers. This thermal e2change is helped b the reciprocating movement of the steam molecules. To simplif & letJs sa that it is the invisible interface between saturated ;and titrated< steam and overheated one which moves at the drum inlet. 42cept in case of incident& the water in its full li5uid state never enters into the drum. ' ver high overheating is possible. Uuus ran non-destro ing tests of coil engines with temperature between 0!! to 1!!NC. *e confirmed with same test on one of our coil engines. 1". Why does the boat 'ibrate# =t is evident that the pulsated water?et contributes to make the boat vibrates& but over all there is a resonance or pendulum effect. 1".1. )elmholt< resonator# #ome ones have searched an analog between pop-pop engine and /elmholt9 resonator. This latter is onl made of a tank ;volume O< prolonged b a pipe of section # and length $. The natural fre5uenc of such a resonator is given b the formulae

$ast updating % Feb.12& 2!1!

Page 21 on 26 c S P c being the velocit of sound in the gas filling the tank. The more classic 2 V/elmholt9 resonator is the bottle which is more or less empt in which air is blown transversall . =t emits a musical sound which depends on the volume of air inside the bottle. =n our application& the tank is the steam drum& the pipe is the pipe and the gas is the steam. 'ccording to most of our e2periments it is true that the fre5uenc evolves with S . :ut the remaining is not so obvious. *hat c and what $ are to be used+ 't low V pressure and relativel low temperature ;that is our case< the steam can be considered as a ...T perfect gas& and c can be calculated b $aplaceJs formula% c = where +mol C" = = 1.)2 BSconstant of the perfect gassesS1)1,>7kmolNC and TStemperature in Cv N-. =t is between )0) and ,!!N- depending upon overheating of the drum. 3mol is the molar mass of the water. =t is the sum of the molar mass of its constituents% /2!S1".666,Q221.!!06S11.!1"2. That gives cY,1"m7s. 3an tests were carried out with a drum ;of ).2cc< provided with a no99le ;of diameter ) and length 11< prolonged b a more or less big and more or less long pipe. The fre5uenc was alwa s the same% appro2 ,.0/9P and et& the strict application of /elmholt9 formula gives fS126!/9. The difference is enormousE The pop-pop engine is definitel not a /elmholt9 resonatorP or if it is& this is invisible because it is high fre5uenc and minute amplitude. f = #ince that time we did some measurements and wrote a small report on that matter. #ee GPop-pop engine and /elmholt9 resonatorI. 1". . Simple resonator. =n mechanics& the simplest resonator needs a fi2ed point where is suspended a mass 3 through a spring of stiffness k. *hen the e5uilibrium is destro ed& this resonator 1 vibrates at the fre5uenc f = . =n practice the 2 + movement is damped due to the friction with the surrounding air. [\ ]

=n h drod namics& the same thing can be done b replacing the spring b a gas and the solid mass b a li5uid one. The most available gas a^_`a\bc\_`d]e`d\]`d^_`fdge] h`ijke`d\]`g]dlle_`\^`d`l\le `f\gm`d`na\^_`e^_`d^_`ojke`f dge]`neajf`ocpp\qeoh`rme`_dk l\^s`\o`kj]e`eppeqg\te`gmd^` \^`gme`l]et\jco`qdoe`neqdcoe `jp`gme`p]\qg\j^`ajooeo`jp`g me`fdge]`\^o\_e`gme`l\leh 1".". %wo-mass resonator.

$ast updating % Feb.12& 2!1!

Page 22 on 26 The fi2ed point of the previous resonators is a sort of infinite mass. *hen no fi2ed point is available we have a two mass resonator. From the mechanical point of view a pop-pop engine can be considered as two masses linked b a spring. .ne of the masses ;31< is the boat ;)!g in our e2ample< and the other one is the mean water mass ;32< inside the pipe ;appro2 1."g<. The spring is the steam.

Z u

Zv

=n such a s stem& when 31 moves of a distance d1 toward left side& 32 moves right +1 b d 2 = d1 . Therefore& the amplitude of the vibrations of the water inside the pipe is +2 5uite more important than the ;visible< one of the boat. =n our e2ample the ratio is about 2!. This allows to GseeI what happens inside the pipe. For that& one eas wa is to fit at the end of the pipe an elbow oriented downward ;or& better& 2 elbows oriented sidewa s if the engine has two pipes<. Thus& the water?et no longer propels the boat. : observing ;with a movie camera because of the relativel high fre5uenc < the boat vibrations we can calculate the amplitude of the water movements inside the pipe. The apparent stiffness of the spring made of steam can be calculated because the fre5uenc is known ;1/9 in our e2ample< thanks to the formula of such a pendulum% 1 ; + 1 + + 2< For our e2ample& this gives a stiffness of appro2 2&"87m. f = 2 + 1.+ 2 Bemarks% #uch an approach could allow to compare pop-pop engines& or to evaluate the performances of an engine depending on ad?ustments. 'ir mi2ed with steam can& depending on the volume& lead to a fre5uenc decrease or increase. 3easurements done later showed that it was mainl the water snake length which was influencing the fre5uenc . 't the first glance& the pop-pop engine looks like an ideal resonator because there is no damping. :ut in fact it is much more complicated than a simple resonator because there is a power suppl from the heat source. -nowing the mean pressure ;1 atmosphere D mean gh of the free surface of the water inside the pipe<& the stiffness& the surface of the cross section of the pipe& and the stroke ;2d< one can calculate the lowest pressure ;"piston" at bottom dead center< d and the highest one ;"piston" at top dead center<. ,ma2 = ,atm + $h + and S d ,min = ,atm + $h . For our to & this gives P ma2T1!).1!)Pa et PminR60.1!)Pa P S that is to sa peak to peak variations of less than 6!!!PaS6!mbar& or about 6!!mmC4. 1".$. =hange o'er from one to the other one. ' stationar pop-pop engine is a one mass resonator. :ut if the engine is light and if ou fit it onboard a ver light hull it becomes a two mass resonator. This same engine could work easil in one case and not in the other one. /ereafter is a message from #later that = received on 1271!72!!0# /0ean1(ves2 I could not $et the sim"le foam boats to %or #

$ast updating % Feb.12& 2!1!

Page 2) on 26 The en$ines seemed to be "erfect but they %ould not start# Finally* I found I could ma e them start by holdin$ onto them firmly# Then it became clear the boats %ould %or if I %ei$hted the foam %ith nails#3 1.".(. =an we use the energy of a resonator# The answer is no& but letJs look wh E $etJs e2amine the e2ample of a grand-father clock. The pendulum is heav and its swings much. To keep going its movement it suffices to bring back up one weight once a week. The corresponding energ is weak because it is used onl to compensate the friction losses. .ne could wish to use the big movements of the heav pendulum@& but it is not possible because the average of this energ is nil. Taking off a few ;one fourth of the stroke< would suffice to stop ever thing. The whole available energ is the potential energ between high and low location of the pendulum. 4S3gh. This energ is weak because h is ver short. 1$. What ;ind of flow in the pipe# VD is to be calculated. =n this formula O is in m7s& ( in m and is the kinetic viscosit in m27s. =f ReT1!!! the flow is laminar. PoiseuilleJs law applies. The pressure losses are proportional to the flow and viscosit & and inversel proportional to the power , of the diameter. The inner roughness has no influence. The linear losses coefficient 6, ;per meter of pipe< is given b = e =f 1!!!TReT2"!! the flow is undetermined. .ften not stable between laminar and turbulent. =f ReR2"!! the flow is turbulent. The pressure drop is appro2imatel proportional to the s5uare of the flow& to the power !.2" of the viscosit and inversel proportional to the power " of the diameter. =f the inner of the pipe is smooth& the !&)16 linear losses coefficient is appro2 = e !&2" For water& the kinetic viscosit in m 27s ;1m27sS1!6 centi#tokes< is about 1.1!-6 at 2!NC& !.66.1!-6 at ,!NC& !.,1.1!-6 at 6!NC& !.)0.1!-6 at 1!NC and !.)!.1!-6 at 1!!NC. For our e2amples& with a water temperature ?ust above 2!NC we will use S!.6.1!-6. For our to & with a ,!mm amplitude& the ma2imum velocit inside the pipe is 1m7s. Corresponding Be nolds number is ReS)660 P hence higher than 2"!!. :ut& as the movement is sinusoidal Re oscillates between ! and )660. For the protot pe nN1 with a 6mm pipe diameter we have noted an amplitude of "!mm and a fre5uenc of ,/9 P this gives a ma2imum velocit of !.62"m7s. Corresponding Be nolds number is ReS,166P hence higher than 2"!!. :ut& as the movement is sinusoidal Re oscillates between ! and ,166. These two e2amples lead to Be nolds numbers that are ver close. Therefore& to optimi9e a pop-pop engine& care will have to be taken as for a turbulent flowP i.e. the roughness of the inside of the pipe will have to be as low as possible. For the protot pe nN2 with a )mm pipe diameter we have noted an amplitude of 2!mm and a fre5uenc of ).)/9 P this gives a ma2imum velocit of !.1m7s. Corresponding Be nolds number is ReS0!!P hence lower than 1!!!. Theoreticall the flow is laminar& To know it& Be nolds number e =

$ast updating % Feb.12& 2!1!

Page 2, on 26 but with the alternative movement up to which level+ =n practice& when in its upper half the pipe is hot where there is water inside& the flow is probabl laminar. 'nd when the pipe is cold it is turbulent. 1(. %he ideal pop-pop engine# The 5uestion mark is intentional. 't the origin ;when we wrote the first release of this document< we had some a priori. 8ow we have got some certainties but there are some shadowed areas. /ow could be the ideal pop-pop engine from the efficienc point of view+ .nl one pipe ;in order to minimi9e the friction losses<. =n fact it is rather eas to fill with a s ringe and a thin fle2ible hose. #mooth inner surface of the pipe. (rum made of thick copper. ;42cellent thermal conductivit % )6!*7mN-< 8o diaphragm. Pipe made of stainless steel. ;Thermal conductivit far less than the one of copper% 1"*7mN-< Thermal shield between fire place and pipe Pipe having a slight and long counter-slope ;so that the dissolved gasses in water evacuate mostl out of the drum@ though we have never met this problem. < #mall volume O of the drum ;to get high pressure variations. < Thick wall and wide area of the lower part of the drum ;1!! times #< to ease instantaneous steam flashing. V #urface of the pipe cross section so that S with # in mm2 and O in mm). 1!! #hape of the no99le% #ee G#haping a no99le for a pop-pop engineI Belationship between pipe length and cross section si9e+ 'nd last& the ideal application would use two identical engines side b side& working in opposite phases to cancel the boat vibrations. =tJs a dream. .n a so simple engine there is nothing that could allow doing such a controlP furthermore& it would be against nature because the longitudinal movements of the boat involved b one of the engine can onl help the phasing of the other one. /owever& on a single engine we have demonstrated that it could be interesting to turn half of the tubes towards the bow and to add a 11!N bend at the end. /ence& the hull vibrations are far less. P.# % =f we e2cept some corrections& practicall all this was written in Februar 2!!" before contacting amateurs and7or specialists of pop-pop engines. There are still things to clarif & and some discrepancies subsist between the points of view of some passionate people. Though man tests demonstrate that a diaphragm decreases the efficienc there are still some people who think it is needed. *e are in favor of the diaphragm for the fun and for its contribution to ease the vibration of the engine. #ome think that some air in the drum is useful. There are Gdr I drums ;most of ours< and drums with alwa s some water in. #ome engines refuse to start or donJt run more than some seconds& and we donJt understand wh . #ometimes it has been noted big discrepancies in the performances though the tests conditions were a priori similar.

$ast updating % Feb.12& 2!1!

Page 2" on 26

'ppro2imatel " ears and 2!! files later@the topic is not closed.

$ast updating % Feb.12& 2!1!

Page 26 on 26 'nne2 1

#ound generator characteristics.

To evaluate the pressures used to deform the membrane the following arrangement was used& using compressed air@ from the mouth of the operator. :lowing or sucking

wjgga e`jp` fdge]

r]d^ol d]e^g` l\le

Besults% #low pressure increase. For hS66mm the membrane becomes suddenl conve2 and emit a pop. For higher pressures nothing. #low release of the pressure. For hS)6 the membrane emits another pop& weaker& and comes back to its original shape nearl flat. #low vacuum increase. For hS-2! the membrane becomes concave and emits again a pop. *ith better vacuum nothing happens. #low release of the vacuum and slow pressure increase. The membrane modifies slightl its shape& but without sound. =t remains slightl concave& and then suddenl becomes conve2 for hS66mm. Therefore& there is a sort of classical h steresis in one wa & and two in the other wa . This is ver likel not made on purpose. =t is a conse5uence of the hand made work with slight strains at rest due to soft metal welding. *hen the drum is submitted 5uickl alternatel to pressure and vacuum& one can hear the t pical pop-pop sound& that is to sa one clear pop in each wa .

$ast updating % Feb.12& 2!1!